diff --git "a/gr00t/run_3real/checkpoint-10000/trainer_state.json" "b/gr00t/run_3real/checkpoint-10000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/gr00t/run_3real/checkpoint-10000/trainer_state.json" @@ -0,0 +1,6034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 1.4688985347747803, + "learning_rate": 1.8e-06, + "loss": 1.38, + "step": 10 + }, + { + "grad_norm": 0.7645030617713928, + "learning_rate": 3.8e-06, + "loss": 1.3564, + "step": 20 + }, + { + "grad_norm": 0.7033658623695374, + "learning_rate": 5.8e-06, + "loss": 1.3255, + "step": 30 + }, + { + "grad_norm": 0.8681603074073792, + "learning_rate": 7.8e-06, + "loss": 1.281, + "step": 40 + }, + { + "grad_norm": 0.7785395383834839, + "learning_rate": 9.800000000000001e-06, + "loss": 1.2172, + "step": 50 + }, + { + "grad_norm": 1.5549288988113403, + "learning_rate": 1.18e-05, + "loss": 1.1603, + "step": 60 + }, + { + "grad_norm": 2.6673245429992676, + "learning_rate": 1.3800000000000002e-05, + "loss": 1.122, + "step": 70 + }, + { + "grad_norm": 1.3578985929489136, + "learning_rate": 1.58e-05, + "loss": 1.0868, + "step": 80 + }, + { + "grad_norm": 1.5027332305908203, + "learning_rate": 1.78e-05, + "loss": 1.0755, + "step": 90 + }, + { + "grad_norm": 1.1409287452697754, + "learning_rate": 1.9800000000000004e-05, + "loss": 1.0629, + "step": 100 + }, + { + "grad_norm": 1.808100938796997, + "learning_rate": 2.18e-05, + "loss": 1.0585, + "step": 110 + }, + { + "grad_norm": 0.5617800951004028, + "learning_rate": 2.38e-05, + "loss": 1.0468, + "step": 120 + }, + { + "grad_norm": 0.7406872510910034, + "learning_rate": 2.58e-05, + "loss": 1.0306, + "step": 130 + }, + { + "grad_norm": 0.3619387745857239, + "learning_rate": 2.7800000000000005e-05, + "loss": 1.0304, + "step": 140 + }, + { + "grad_norm": 0.784307062625885, + "learning_rate": 2.98e-05, + "loss": 1.0406, + "step": 150 + }, + { + "grad_norm": 0.3664414882659912, + "learning_rate": 3.18e-05, + "loss": 1.0316, + "step": 160 + }, + { + "grad_norm": 0.4572850465774536, + "learning_rate": 3.38e-05, + "loss": 1.0263, + "step": 170 + }, + { + "grad_norm": 0.3709733486175537, + "learning_rate": 3.58e-05, + "loss": 1.0298, + "step": 180 + }, + { + "grad_norm": 0.4023006856441498, + "learning_rate": 3.7800000000000004e-05, + "loss": 1.0219, + "step": 190 + }, + { + "grad_norm": 0.35605332255363464, + "learning_rate": 3.9800000000000005e-05, + "loss": 1.0251, + "step": 200 + }, + { + "grad_norm": 0.6693720817565918, + "learning_rate": 4.18e-05, + "loss": 1.0224, + "step": 210 + }, + { + "grad_norm": 0.5391326546669006, + "learning_rate": 4.38e-05, + "loss": 1.0199, + "step": 220 + }, + { + "grad_norm": 0.6245490908622742, + "learning_rate": 4.58e-05, + "loss": 1.0064, + "step": 230 + }, + { + "grad_norm": 0.5257537364959717, + "learning_rate": 4.78e-05, + "loss": 0.9864, + "step": 240 + }, + { + "grad_norm": 0.49752286076545715, + "learning_rate": 4.9800000000000004e-05, + "loss": 0.9742, + "step": 250 + }, + { + "grad_norm": 0.7133261561393738, + "learning_rate": 5.1800000000000005e-05, + "loss": 0.9543, + "step": 260 + }, + { + "grad_norm": 0.6734635233879089, + "learning_rate": 5.380000000000001e-05, + "loss": 0.9273, + "step": 270 + }, + { + "grad_norm": 0.6387033462524414, + "learning_rate": 5.580000000000001e-05, + "loss": 0.897, + "step": 280 + }, + { + "grad_norm": 0.6541304588317871, + "learning_rate": 5.7799999999999995e-05, + "loss": 0.8742, + "step": 290 + }, + { + "grad_norm": 0.6810693740844727, + "learning_rate": 5.9800000000000003e-05, + "loss": 0.8282, + "step": 300 + }, + { + "grad_norm": 0.848870575428009, + "learning_rate": 6.18e-05, + "loss": 0.7891, + "step": 310 + }, + { + "grad_norm": 0.8962128758430481, + "learning_rate": 6.38e-05, + "loss": 0.745, + "step": 320 + }, + { + "grad_norm": 0.8951132893562317, + "learning_rate": 6.58e-05, + "loss": 0.7059, + "step": 330 + }, + { + "grad_norm": 0.7835540175437927, + "learning_rate": 6.780000000000001e-05, + "loss": 0.6712, + "step": 340 + }, + { + "grad_norm": 0.8961076736450195, + "learning_rate": 6.98e-05, + "loss": 0.6404, + "step": 350 + }, + { + "grad_norm": 0.7485954165458679, + "learning_rate": 7.18e-05, + "loss": 0.6196, + "step": 360 + }, + { + "grad_norm": 0.7506619691848755, + "learning_rate": 7.38e-05, + "loss": 0.6047, + "step": 370 + }, + { + "grad_norm": 0.8093503713607788, + "learning_rate": 7.58e-05, + "loss": 0.5747, + "step": 380 + }, + { + "grad_norm": 0.7800814509391785, + "learning_rate": 7.780000000000001e-05, + "loss": 0.5642, + "step": 390 + }, + { + "grad_norm": 0.8147926330566406, + "learning_rate": 7.98e-05, + "loss": 0.543, + "step": 400 + }, + { + "grad_norm": 0.9376605749130249, + "learning_rate": 8.18e-05, + "loss": 0.5307, + "step": 410 + }, + { + "grad_norm": 0.9530451893806458, + "learning_rate": 8.38e-05, + "loss": 0.4931, + "step": 420 + }, + { + "grad_norm": 1.0562382936477661, + "learning_rate": 8.58e-05, + "loss": 0.4746, + "step": 430 + }, + { + "grad_norm": 0.9154590964317322, + "learning_rate": 8.78e-05, + "loss": 0.4608, + "step": 440 + }, + { + "grad_norm": 0.798561692237854, + "learning_rate": 8.98e-05, + "loss": 0.4455, + "step": 450 + }, + { + "grad_norm": 0.8828734755516052, + "learning_rate": 9.180000000000001e-05, + "loss": 0.448, + "step": 460 + }, + { + "grad_norm": 0.8360201120376587, + "learning_rate": 9.38e-05, + "loss": 0.4424, + "step": 470 + }, + { + "grad_norm": 0.6740878224372864, + "learning_rate": 9.58e-05, + "loss": 0.4293, + "step": 480 + }, + { + "grad_norm": 0.7638790011405945, + "learning_rate": 9.78e-05, + "loss": 0.4262, + "step": 490 + }, + { + "grad_norm": 0.6575111150741577, + "learning_rate": 9.98e-05, + "loss": 0.4145, + "step": 500 + }, + { + "grad_norm": 0.7660487294197083, + "learning_rate": 9.9999778549206e-05, + "loss": 0.4074, + "step": 510 + }, + { + "grad_norm": 0.7423355579376221, + "learning_rate": 9.999901304280685e-05, + "loss": 0.3972, + "step": 520 + }, + { + "grad_norm": 0.9097589254379272, + "learning_rate": 9.999770075521164e-05, + "loss": 0.3977, + "step": 530 + }, + { + "grad_norm": 0.8944575190544128, + "learning_rate": 9.99958417007713e-05, + "loss": 0.3812, + "step": 540 + }, + { + "grad_norm": 0.956577718257904, + "learning_rate": 9.999343589981615e-05, + "loss": 0.3793, + "step": 550 + }, + { + "grad_norm": 0.8228615522384644, + "learning_rate": 9.999048337865568e-05, + "loss": 0.3581, + "step": 560 + }, + { + "grad_norm": 0.7517023682594299, + "learning_rate": 9.998698416957815e-05, + "loss": 0.3562, + "step": 570 + }, + { + "grad_norm": 0.7517921328544617, + "learning_rate": 9.998293831085037e-05, + "loss": 0.3557, + "step": 580 + }, + { + "grad_norm": 0.7406057715415955, + "learning_rate": 9.997834584671719e-05, + "loss": 0.3525, + "step": 590 + }, + { + "grad_norm": 0.6713485717773438, + "learning_rate": 9.997320682740107e-05, + "loss": 0.3493, + "step": 600 + }, + { + "grad_norm": 0.9752911925315857, + "learning_rate": 9.996752130910149e-05, + "loss": 0.3464, + "step": 610 + }, + { + "grad_norm": 0.8473848104476929, + "learning_rate": 9.99612893539944e-05, + "loss": 0.3317, + "step": 620 + }, + { + "grad_norm": 0.717465877532959, + "learning_rate": 9.995451103023144e-05, + "loss": 0.3292, + "step": 630 + }, + { + "grad_norm": 0.8652439713478088, + "learning_rate": 9.994718641193928e-05, + "loss": 0.3323, + "step": 640 + }, + { + "grad_norm": 0.7446794509887695, + "learning_rate": 9.993931557921874e-05, + "loss": 0.3258, + "step": 650 + }, + { + "grad_norm": 0.7172669768333435, + "learning_rate": 9.993089861814402e-05, + "loss": 0.3258, + "step": 660 + }, + { + "grad_norm": 0.669628381729126, + "learning_rate": 9.992193562076166e-05, + "loss": 0.3377, + "step": 670 + }, + { + "grad_norm": 0.8395833373069763, + "learning_rate": 9.991242668508954e-05, + "loss": 0.3227, + "step": 680 + }, + { + "grad_norm": 0.7585325241088867, + "learning_rate": 9.990237191511587e-05, + "loss": 0.3227, + "step": 690 + }, + { + "grad_norm": 0.6546490788459778, + "learning_rate": 9.989177142079802e-05, + "loss": 0.3224, + "step": 700 + }, + { + "grad_norm": 0.626818835735321, + "learning_rate": 9.988062531806126e-05, + "loss": 0.3277, + "step": 710 + }, + { + "grad_norm": 0.6902979612350464, + "learning_rate": 9.986893372879762e-05, + "loss": 0.3155, + "step": 720 + }, + { + "grad_norm": 0.6836565732955933, + "learning_rate": 9.985669678086443e-05, + "loss": 0.3179, + "step": 730 + }, + { + "grad_norm": 0.7668564915657043, + "learning_rate": 9.984391460808298e-05, + "loss": 0.3061, + "step": 740 + }, + { + "grad_norm": 0.7735840082168579, + "learning_rate": 9.983058735023709e-05, + "loss": 0.3148, + "step": 750 + }, + { + "grad_norm": 0.7396438717842102, + "learning_rate": 9.98167151530715e-05, + "loss": 0.2984, + "step": 760 + }, + { + "grad_norm": 0.7460150718688965, + "learning_rate": 9.980229816829034e-05, + "loss": 0.2978, + "step": 770 + }, + { + "grad_norm": 0.7281282544136047, + "learning_rate": 9.978733655355544e-05, + "loss": 0.2983, + "step": 780 + }, + { + "grad_norm": 0.7528439164161682, + "learning_rate": 9.977183047248464e-05, + "loss": 0.2973, + "step": 790 + }, + { + "grad_norm": 0.7017455101013184, + "learning_rate": 9.975578009464992e-05, + "loss": 0.3008, + "step": 800 + }, + { + "grad_norm": 0.7145857810974121, + "learning_rate": 9.97391855955757e-05, + "loss": 0.3006, + "step": 810 + }, + { + "grad_norm": 0.671181857585907, + "learning_rate": 9.972204715673669e-05, + "loss": 0.292, + "step": 820 + }, + { + "grad_norm": 0.8692606687545776, + "learning_rate": 9.970436496555617e-05, + "loss": 0.2913, + "step": 830 + }, + { + "grad_norm": 0.9781755805015564, + "learning_rate": 9.968613921540373e-05, + "loss": 0.2849, + "step": 840 + }, + { + "grad_norm": 0.906934916973114, + "learning_rate": 9.966737010559326e-05, + "loss": 0.2876, + "step": 850 + }, + { + "grad_norm": 0.8187403082847595, + "learning_rate": 9.964805784138072e-05, + "loss": 0.2766, + "step": 860 + }, + { + "grad_norm": 0.6051341891288757, + "learning_rate": 9.962820263396195e-05, + "loss": 0.2632, + "step": 870 + }, + { + "grad_norm": 0.7972335815429688, + "learning_rate": 9.960780470047033e-05, + "loss": 0.2577, + "step": 880 + }, + { + "grad_norm": 0.6247836351394653, + "learning_rate": 9.958686426397437e-05, + "loss": 0.2583, + "step": 890 + }, + { + "grad_norm": 0.7198521494865417, + "learning_rate": 9.956538155347534e-05, + "loss": 0.2679, + "step": 900 + }, + { + "grad_norm": 0.6863376498222351, + "learning_rate": 9.95433568039047e-05, + "loss": 0.2467, + "step": 910 + }, + { + "grad_norm": 0.6362172365188599, + "learning_rate": 9.952079025612162e-05, + "loss": 0.2514, + "step": 920 + }, + { + "grad_norm": 0.6610420942306519, + "learning_rate": 9.949768215691022e-05, + "loss": 0.2464, + "step": 930 + }, + { + "grad_norm": 0.735052227973938, + "learning_rate": 9.9474032758977e-05, + "loss": 0.2432, + "step": 940 + }, + { + "grad_norm": 0.5371904969215393, + "learning_rate": 9.944984232094794e-05, + "loss": 0.2507, + "step": 950 + }, + { + "grad_norm": 0.6107508540153503, + "learning_rate": 9.942511110736584e-05, + "loss": 0.2363, + "step": 960 + }, + { + "grad_norm": 0.7737448811531067, + "learning_rate": 9.939983938868726e-05, + "loss": 0.2331, + "step": 970 + }, + { + "grad_norm": 0.7212536931037903, + "learning_rate": 9.93740274412797e-05, + "loss": 0.2315, + "step": 980 + }, + { + "grad_norm": 0.5824896097183228, + "learning_rate": 9.934767554741846e-05, + "loss": 0.2315, + "step": 990 + }, + { + "grad_norm": 0.6824327111244202, + "learning_rate": 9.932078399528361e-05, + "loss": 0.225, + "step": 1000 + }, + { + "grad_norm": 0.6186544895172119, + "learning_rate": 9.929335307895689e-05, + "loss": 0.2206, + "step": 1010 + }, + { + "grad_norm": 0.7479929327964783, + "learning_rate": 9.926538309841839e-05, + "loss": 0.221, + "step": 1020 + }, + { + "grad_norm": 0.5823713541030884, + "learning_rate": 9.923687435954334e-05, + "loss": 0.222, + "step": 1030 + }, + { + "grad_norm": 0.7993959188461304, + "learning_rate": 9.920782717409873e-05, + "loss": 0.2125, + "step": 1040 + }, + { + "grad_norm": 0.5921409130096436, + "learning_rate": 9.917824185973994e-05, + "loss": 0.2059, + "step": 1050 + }, + { + "grad_norm": 0.8483374714851379, + "learning_rate": 9.914811874000723e-05, + "loss": 0.2142, + "step": 1060 + }, + { + "grad_norm": 0.6285680532455444, + "learning_rate": 9.911745814432218e-05, + "loss": 0.2116, + "step": 1070 + }, + { + "grad_norm": 0.6570925116539001, + "learning_rate": 9.90862604079842e-05, + "loss": 0.2109, + "step": 1080 + }, + { + "grad_norm": 0.6355977058410645, + "learning_rate": 9.90545258721667e-05, + "loss": 0.2015, + "step": 1090 + }, + { + "grad_norm": 0.8277871608734131, + "learning_rate": 9.90222548839135e-05, + "loss": 0.2005, + "step": 1100 + }, + { + "grad_norm": 0.642349123954773, + "learning_rate": 9.898944779613495e-05, + "loss": 0.2038, + "step": 1110 + }, + { + "grad_norm": 0.6685327887535095, + "learning_rate": 9.89561049676041e-05, + "loss": 0.2015, + "step": 1120 + }, + { + "grad_norm": 0.7061478495597839, + "learning_rate": 9.89222267629528e-05, + "loss": 0.2081, + "step": 1130 + }, + { + "grad_norm": 0.7631548047065735, + "learning_rate": 9.888781355266763e-05, + "loss": 0.2118, + "step": 1140 + }, + { + "grad_norm": 0.651374101638794, + "learning_rate": 9.885286571308598e-05, + "loss": 0.2075, + "step": 1150 + }, + { + "grad_norm": 0.6582494378089905, + "learning_rate": 9.881738362639182e-05, + "loss": 0.2049, + "step": 1160 + }, + { + "grad_norm": 0.7720977663993835, + "learning_rate": 9.878136768061154e-05, + "loss": 0.2075, + "step": 1170 + }, + { + "grad_norm": 0.6204569935798645, + "learning_rate": 9.874481826960979e-05, + "loss": 0.2054, + "step": 1180 + }, + { + "grad_norm": 0.7273457646369934, + "learning_rate": 9.870773579308503e-05, + "loss": 0.2084, + "step": 1190 + }, + { + "grad_norm": 0.500053882598877, + "learning_rate": 9.867012065656533e-05, + "loss": 0.2056, + "step": 1200 + }, + { + "grad_norm": 0.6381528973579407, + "learning_rate": 9.863197327140376e-05, + "loss": 0.2006, + "step": 1210 + }, + { + "grad_norm": 0.6237829327583313, + "learning_rate": 9.859329405477403e-05, + "loss": 0.2001, + "step": 1220 + }, + { + "grad_norm": 0.6929218769073486, + "learning_rate": 9.855408342966585e-05, + "loss": 0.1998, + "step": 1230 + }, + { + "grad_norm": 0.6542112827301025, + "learning_rate": 9.851434182488033e-05, + "loss": 0.1958, + "step": 1240 + }, + { + "grad_norm": 0.7060437798500061, + "learning_rate": 9.84740696750253e-05, + "loss": 0.201, + "step": 1250 + }, + { + "grad_norm": 0.7238238453865051, + "learning_rate": 9.843326742051055e-05, + "loss": 0.1949, + "step": 1260 + }, + { + "grad_norm": 0.7471667528152466, + "learning_rate": 9.839193550754297e-05, + "loss": 0.2044, + "step": 1270 + }, + { + "grad_norm": 0.5196937322616577, + "learning_rate": 9.835007438812177e-05, + "loss": 0.198, + "step": 1280 + }, + { + "grad_norm": 0.5683590173721313, + "learning_rate": 9.830768452003341e-05, + "loss": 0.1984, + "step": 1290 + }, + { + "grad_norm": 0.8583512902259827, + "learning_rate": 9.826476636684671e-05, + "loss": 0.2053, + "step": 1300 + }, + { + "grad_norm": 0.7557379007339478, + "learning_rate": 9.822132039790773e-05, + "loss": 0.1986, + "step": 1310 + }, + { + "grad_norm": 0.6582274436950684, + "learning_rate": 9.817734708833461e-05, + "loss": 0.202, + "step": 1320 + }, + { + "grad_norm": 0.7102360725402832, + "learning_rate": 9.813284691901243e-05, + "loss": 0.1934, + "step": 1330 + }, + { + "grad_norm": 0.6746431589126587, + "learning_rate": 9.808782037658792e-05, + "loss": 0.1945, + "step": 1340 + }, + { + "grad_norm": 0.5591486096382141, + "learning_rate": 9.804226795346411e-05, + "loss": 0.1834, + "step": 1350 + }, + { + "grad_norm": 0.893364667892456, + "learning_rate": 9.799619014779503e-05, + "loss": 0.193, + "step": 1360 + }, + { + "grad_norm": 0.6140760183334351, + "learning_rate": 9.794958746348013e-05, + "loss": 0.1782, + "step": 1370 + }, + { + "grad_norm": 0.6883726716041565, + "learning_rate": 9.790246041015896e-05, + "loss": 0.1818, + "step": 1380 + }, + { + "grad_norm": 0.789801299571991, + "learning_rate": 9.785480950320538e-05, + "loss": 0.1761, + "step": 1390 + }, + { + "grad_norm": 0.7107528448104858, + "learning_rate": 9.78066352637221e-05, + "loss": 0.1798, + "step": 1400 + }, + { + "grad_norm": 0.6448025107383728, + "learning_rate": 9.775793821853488e-05, + "loss": 0.1659, + "step": 1410 + }, + { + "grad_norm": 0.6221340298652649, + "learning_rate": 9.77087189001868e-05, + "loss": 0.1795, + "step": 1420 + }, + { + "grad_norm": 0.6247802972793579, + "learning_rate": 9.765897784693243e-05, + "loss": 0.1714, + "step": 1430 + }, + { + "grad_norm": 0.6431756615638733, + "learning_rate": 9.760871560273197e-05, + "loss": 0.1719, + "step": 1440 + }, + { + "grad_norm": 0.6085208654403687, + "learning_rate": 9.755793271724526e-05, + "loss": 0.1793, + "step": 1450 + }, + { + "grad_norm": 0.5787495374679565, + "learning_rate": 9.750662974582584e-05, + "loss": 0.1733, + "step": 1460 + }, + { + "grad_norm": 0.5149695873260498, + "learning_rate": 9.745480724951473e-05, + "loss": 0.1711, + "step": 1470 + }, + { + "grad_norm": 0.5747590661048889, + "learning_rate": 9.740246579503447e-05, + "loss": 0.1703, + "step": 1480 + }, + { + "grad_norm": 0.5937369465827942, + "learning_rate": 9.734960595478284e-05, + "loss": 0.1651, + "step": 1490 + }, + { + "grad_norm": 0.758808970451355, + "learning_rate": 9.729622830682657e-05, + "loss": 0.1763, + "step": 1500 + }, + { + "grad_norm": 0.8273259997367859, + "learning_rate": 9.724233343489504e-05, + "loss": 0.1631, + "step": 1510 + }, + { + "grad_norm": 0.6177854537963867, + "learning_rate": 9.718792192837396e-05, + "loss": 0.1563, + "step": 1520 + }, + { + "grad_norm": 0.7404709458351135, + "learning_rate": 9.713299438229886e-05, + "loss": 0.1581, + "step": 1530 + }, + { + "grad_norm": 0.6295482516288757, + "learning_rate": 9.707755139734855e-05, + "loss": 0.1642, + "step": 1540 + }, + { + "grad_norm": 0.6740186214447021, + "learning_rate": 9.702159357983866e-05, + "loss": 0.1542, + "step": 1550 + }, + { + "grad_norm": 0.6013017296791077, + "learning_rate": 9.696512154171492e-05, + "loss": 0.1501, + "step": 1560 + }, + { + "grad_norm": 0.5249996781349182, + "learning_rate": 9.690813590054645e-05, + "loss": 0.1473, + "step": 1570 + }, + { + "grad_norm": 0.696631133556366, + "learning_rate": 9.685063727951914e-05, + "loss": 0.1482, + "step": 1580 + }, + { + "grad_norm": 0.589519202709198, + "learning_rate": 9.679262630742865e-05, + "loss": 0.149, + "step": 1590 + }, + { + "grad_norm": 0.6700538992881775, + "learning_rate": 9.673410361867373e-05, + "loss": 0.151, + "step": 1600 + }, + { + "grad_norm": 0.5307331085205078, + "learning_rate": 9.667506985324909e-05, + "loss": 0.1433, + "step": 1610 + }, + { + "grad_norm": 0.5697365403175354, + "learning_rate": 9.661552565673855e-05, + "loss": 0.1456, + "step": 1620 + }, + { + "grad_norm": 0.5164391994476318, + "learning_rate": 9.655547168030789e-05, + "loss": 0.1478, + "step": 1630 + }, + { + "grad_norm": 0.6379405856132507, + "learning_rate": 9.649490858069777e-05, + "loss": 0.1472, + "step": 1640 + }, + { + "grad_norm": 0.5417969822883606, + "learning_rate": 9.643383702021658e-05, + "loss": 0.1496, + "step": 1650 + }, + { + "grad_norm": 0.628899097442627, + "learning_rate": 9.637225766673307e-05, + "loss": 0.1398, + "step": 1660 + }, + { + "grad_norm": 0.6133635640144348, + "learning_rate": 9.631017119366922e-05, + "loss": 0.1483, + "step": 1670 + }, + { + "grad_norm": 0.5963024497032166, + "learning_rate": 9.624757827999273e-05, + "loss": 0.1476, + "step": 1680 + }, + { + "grad_norm": 0.6431703567504883, + "learning_rate": 9.618447961020971e-05, + "loss": 0.147, + "step": 1690 + }, + { + "grad_norm": 0.6507049798965454, + "learning_rate": 9.612087587435707e-05, + "loss": 0.1403, + "step": 1700 + }, + { + "grad_norm": 0.7144182324409485, + "learning_rate": 9.605676776799508e-05, + "loss": 0.1364, + "step": 1710 + }, + { + "grad_norm": 0.7109726667404175, + "learning_rate": 9.599215599219973e-05, + "loss": 0.1384, + "step": 1720 + }, + { + "grad_norm": 0.6847316026687622, + "learning_rate": 9.592704125355505e-05, + "loss": 0.1317, + "step": 1730 + }, + { + "grad_norm": 0.6259592771530151, + "learning_rate": 9.586142426414538e-05, + "loss": 0.1222, + "step": 1740 + }, + { + "grad_norm": 0.8183260560035706, + "learning_rate": 9.57953057415476e-05, + "loss": 0.131, + "step": 1750 + }, + { + "grad_norm": 0.7058231234550476, + "learning_rate": 9.572868640882328e-05, + "loss": 0.1322, + "step": 1760 + }, + { + "grad_norm": 0.6913846135139465, + "learning_rate": 9.56615669945108e-05, + "loss": 0.1267, + "step": 1770 + }, + { + "grad_norm": 0.5570674538612366, + "learning_rate": 9.55939482326173e-05, + "loss": 0.122, + "step": 1780 + }, + { + "grad_norm": 0.7559832334518433, + "learning_rate": 9.552583086261069e-05, + "loss": 0.1271, + "step": 1790 + }, + { + "grad_norm": 0.7949308156967163, + "learning_rate": 9.545721562941168e-05, + "loss": 0.1144, + "step": 1800 + }, + { + "grad_norm": 0.6073559522628784, + "learning_rate": 9.538810328338543e-05, + "loss": 0.1081, + "step": 1810 + }, + { + "grad_norm": 0.6428042054176331, + "learning_rate": 9.531849458033349e-05, + "loss": 0.1019, + "step": 1820 + }, + { + "grad_norm": 0.8170904517173767, + "learning_rate": 9.524839028148547e-05, + "loss": 0.104, + "step": 1830 + }, + { + "grad_norm": 0.6276879906654358, + "learning_rate": 9.517779115349077e-05, + "loss": 0.0993, + "step": 1840 + }, + { + "grad_norm": 0.6298386454582214, + "learning_rate": 9.510669796841014e-05, + "loss": 0.0999, + "step": 1850 + }, + { + "grad_norm": 0.8434498310089111, + "learning_rate": 9.503511150370727e-05, + "loss": 0.0999, + "step": 1860 + }, + { + "grad_norm": 0.7385769486427307, + "learning_rate": 9.496303254224024e-05, + "loss": 0.098, + "step": 1870 + }, + { + "grad_norm": 0.6802159547805786, + "learning_rate": 9.489046187225306e-05, + "loss": 0.0845, + "step": 1880 + }, + { + "grad_norm": 0.6350604295730591, + "learning_rate": 9.481740028736692e-05, + "loss": 0.0807, + "step": 1890 + }, + { + "grad_norm": 0.7797428965568542, + "learning_rate": 9.474384858657164e-05, + "loss": 0.0996, + "step": 1900 + }, + { + "grad_norm": 0.7418763041496277, + "learning_rate": 9.466980757421679e-05, + "loss": 0.0886, + "step": 1910 + }, + { + "grad_norm": 0.6989834308624268, + "learning_rate": 9.459527806000305e-05, + "loss": 0.0832, + "step": 1920 + }, + { + "grad_norm": 0.6905877590179443, + "learning_rate": 9.452026085897325e-05, + "loss": 0.0891, + "step": 1930 + }, + { + "grad_norm": 0.5993806719779968, + "learning_rate": 9.444475679150348e-05, + "loss": 0.0846, + "step": 1940 + }, + { + "grad_norm": 0.6589913964271545, + "learning_rate": 9.436876668329411e-05, + "loss": 0.087, + "step": 1950 + }, + { + "grad_norm": 0.6733335256576538, + "learning_rate": 9.429229136536079e-05, + "loss": 0.0783, + "step": 1960 + }, + { + "grad_norm": 0.6809325814247131, + "learning_rate": 9.421533167402534e-05, + "loss": 0.0871, + "step": 1970 + }, + { + "grad_norm": 0.8038035035133362, + "learning_rate": 9.413788845090666e-05, + "loss": 0.094, + "step": 1980 + }, + { + "grad_norm": 0.49830174446105957, + "learning_rate": 9.405996254291136e-05, + "loss": 0.0835, + "step": 1990 + }, + { + "grad_norm": 0.5099689960479736, + "learning_rate": 9.398155480222474e-05, + "loss": 0.0795, + "step": 2000 + }, + { + "grad_norm": 0.5958901047706604, + "learning_rate": 9.390266608630128e-05, + "loss": 0.0755, + "step": 2010 + }, + { + "grad_norm": 0.6426071524620056, + "learning_rate": 9.38232972578553e-05, + "loss": 0.082, + "step": 2020 + }, + { + "grad_norm": 0.6249590516090393, + "learning_rate": 9.374344918485164e-05, + "loss": 0.0822, + "step": 2030 + }, + { + "grad_norm": 0.5605111122131348, + "learning_rate": 9.366312274049602e-05, + "loss": 0.0883, + "step": 2040 + }, + { + "grad_norm": 0.7479569315910339, + "learning_rate": 9.358231880322554e-05, + "loss": 0.0803, + "step": 2050 + }, + { + "grad_norm": 0.7081699967384338, + "learning_rate": 9.350103825669916e-05, + "loss": 0.0778, + "step": 2060 + }, + { + "grad_norm": 0.6158730983734131, + "learning_rate": 9.341928198978787e-05, + "loss": 0.08, + "step": 2070 + }, + { + "grad_norm": 0.552537202835083, + "learning_rate": 9.333705089656512e-05, + "loss": 0.0723, + "step": 2080 + }, + { + "grad_norm": 0.6489994525909424, + "learning_rate": 9.325434587629698e-05, + "loss": 0.075, + "step": 2090 + }, + { + "grad_norm": 0.5815051198005676, + "learning_rate": 9.31711678334323e-05, + "loss": 0.0774, + "step": 2100 + }, + { + "grad_norm": 0.4953973591327667, + "learning_rate": 9.308751767759282e-05, + "loss": 0.0747, + "step": 2110 + }, + { + "grad_norm": 0.4764729142189026, + "learning_rate": 9.300339632356325e-05, + "loss": 0.0695, + "step": 2120 + }, + { + "grad_norm": 0.4616457521915436, + "learning_rate": 9.291880469128124e-05, + "loss": 0.0714, + "step": 2130 + }, + { + "grad_norm": 0.5330332517623901, + "learning_rate": 9.283374370582732e-05, + "loss": 0.0754, + "step": 2140 + }, + { + "grad_norm": 0.6691517233848572, + "learning_rate": 9.274821429741482e-05, + "loss": 0.0814, + "step": 2150 + }, + { + "grad_norm": 0.6739680767059326, + "learning_rate": 9.266221740137961e-05, + "loss": 0.0744, + "step": 2160 + }, + { + "grad_norm": 0.5638138055801392, + "learning_rate": 9.257575395817001e-05, + "loss": 0.0715, + "step": 2170 + }, + { + "grad_norm": 0.619041383266449, + "learning_rate": 9.248882491333637e-05, + "loss": 0.0757, + "step": 2180 + }, + { + "grad_norm": 0.6553375124931335, + "learning_rate": 9.240143121752076e-05, + "loss": 0.073, + "step": 2190 + }, + { + "grad_norm": 0.6725075244903564, + "learning_rate": 9.23135738264467e-05, + "loss": 0.0712, + "step": 2200 + }, + { + "grad_norm": 0.5573500394821167, + "learning_rate": 9.222525370090849e-05, + "loss": 0.0693, + "step": 2210 + }, + { + "grad_norm": 0.5087471604347229, + "learning_rate": 9.213647180676088e-05, + "loss": 0.0662, + "step": 2220 + }, + { + "grad_norm": 0.6130621433258057, + "learning_rate": 9.204722911490846e-05, + "loss": 0.0712, + "step": 2230 + }, + { + "grad_norm": 0.5628780722618103, + "learning_rate": 9.1957526601295e-05, + "loss": 0.0592, + "step": 2240 + }, + { + "grad_norm": 0.9200260639190674, + "learning_rate": 9.186736524689281e-05, + "loss": 0.0658, + "step": 2250 + }, + { + "grad_norm": 0.6945054531097412, + "learning_rate": 9.177674603769204e-05, + "loss": 0.0638, + "step": 2260 + }, + { + "grad_norm": 0.8129741549491882, + "learning_rate": 9.168566996468983e-05, + "loss": 0.0601, + "step": 2270 + }, + { + "grad_norm": 0.6894360184669495, + "learning_rate": 9.159413802387951e-05, + "loss": 0.0666, + "step": 2280 + }, + { + "grad_norm": 0.5961335301399231, + "learning_rate": 9.150215121623974e-05, + "loss": 0.0605, + "step": 2290 + }, + { + "grad_norm": 0.7185078263282776, + "learning_rate": 9.140971054772349e-05, + "loss": 0.0547, + "step": 2300 + }, + { + "grad_norm": 0.5971482396125793, + "learning_rate": 9.131681702924713e-05, + "loss": 0.0642, + "step": 2310 + }, + { + "grad_norm": 0.8267772197723389, + "learning_rate": 9.122347167667926e-05, + "loss": 0.0533, + "step": 2320 + }, + { + "grad_norm": 0.7043777108192444, + "learning_rate": 9.112967551082973e-05, + "loss": 0.0649, + "step": 2330 + }, + { + "grad_norm": 0.7251552939414978, + "learning_rate": 9.103542955743835e-05, + "loss": 0.0526, + "step": 2340 + }, + { + "grad_norm": 0.641697108745575, + "learning_rate": 9.094073484716381e-05, + "loss": 0.0605, + "step": 2350 + }, + { + "grad_norm": 0.6640056371688843, + "learning_rate": 9.084559241557226e-05, + "loss": 0.0528, + "step": 2360 + }, + { + "grad_norm": 0.6051256060600281, + "learning_rate": 9.075000330312608e-05, + "loss": 0.0551, + "step": 2370 + }, + { + "grad_norm": 0.6508076190948486, + "learning_rate": 9.065396855517253e-05, + "loss": 0.0533, + "step": 2380 + }, + { + "grad_norm": 0.5256751179695129, + "learning_rate": 9.055748922193219e-05, + "loss": 0.0595, + "step": 2390 + }, + { + "grad_norm": 0.5923941135406494, + "learning_rate": 9.046056635848761e-05, + "loss": 0.0601, + "step": 2400 + }, + { + "grad_norm": 0.6456895470619202, + "learning_rate": 9.036320102477169e-05, + "loss": 0.0505, + "step": 2410 + }, + { + "grad_norm": 0.6933112144470215, + "learning_rate": 9.02653942855561e-05, + "loss": 0.0554, + "step": 2420 + }, + { + "grad_norm": 0.6988584399223328, + "learning_rate": 9.016714721043971e-05, + "loss": 0.0531, + "step": 2430 + }, + { + "grad_norm": 0.5013759136199951, + "learning_rate": 9.006846087383675e-05, + "loss": 0.0515, + "step": 2440 + }, + { + "grad_norm": 0.4983830749988556, + "learning_rate": 8.996933635496523e-05, + "loss": 0.0535, + "step": 2450 + }, + { + "grad_norm": 0.6666116118431091, + "learning_rate": 8.986977473783498e-05, + "loss": 0.0567, + "step": 2460 + }, + { + "grad_norm": 0.5705677270889282, + "learning_rate": 8.97697771112359e-05, + "loss": 0.0529, + "step": 2470 + }, + { + "grad_norm": 0.6851279139518738, + "learning_rate": 8.966934456872602e-05, + "loss": 0.0558, + "step": 2480 + }, + { + "grad_norm": 0.649524450302124, + "learning_rate": 8.95684782086195e-05, + "loss": 0.0492, + "step": 2490 + }, + { + "grad_norm": 0.5539266467094421, + "learning_rate": 8.946717913397476e-05, + "loss": 0.0585, + "step": 2500 + }, + { + "grad_norm": 0.6333678364753723, + "learning_rate": 8.93654484525822e-05, + "loss": 0.053, + "step": 2510 + }, + { + "grad_norm": 0.5263622403144836, + "learning_rate": 8.926328727695226e-05, + "loss": 0.053, + "step": 2520 + }, + { + "grad_norm": 0.6450868844985962, + "learning_rate": 8.916069672430319e-05, + "loss": 0.0512, + "step": 2530 + }, + { + "grad_norm": 0.5581928491592407, + "learning_rate": 8.905767791654884e-05, + "loss": 0.0573, + "step": 2540 + }, + { + "grad_norm": 0.5718982219696045, + "learning_rate": 8.895423198028638e-05, + "loss": 0.0614, + "step": 2550 + }, + { + "grad_norm": 0.6775624752044678, + "learning_rate": 8.885036004678402e-05, + "loss": 0.059, + "step": 2560 + }, + { + "grad_norm": 0.6321313977241516, + "learning_rate": 8.874606325196857e-05, + "loss": 0.0518, + "step": 2570 + }, + { + "grad_norm": 0.6444363594055176, + "learning_rate": 8.864134273641304e-05, + "loss": 0.0528, + "step": 2580 + }, + { + "grad_norm": 0.5716114640235901, + "learning_rate": 8.853619964532427e-05, + "loss": 0.0523, + "step": 2590 + }, + { + "grad_norm": 0.6638785004615784, + "learning_rate": 8.843063512853019e-05, + "loss": 0.0507, + "step": 2600 + }, + { + "grad_norm": 0.6005464196205139, + "learning_rate": 8.832465034046749e-05, + "loss": 0.0524, + "step": 2610 + }, + { + "grad_norm": 0.5212604999542236, + "learning_rate": 8.821824644016882e-05, + "loss": 0.0513, + "step": 2620 + }, + { + "grad_norm": 0.572059690952301, + "learning_rate": 8.811142459125019e-05, + "loss": 0.0504, + "step": 2630 + }, + { + "grad_norm": 0.48632654547691345, + "learning_rate": 8.800418596189822e-05, + "loss": 0.0463, + "step": 2640 + }, + { + "grad_norm": 0.6325863003730774, + "learning_rate": 8.789653172485737e-05, + "loss": 0.0517, + "step": 2650 + }, + { + "grad_norm": 0.637148380279541, + "learning_rate": 8.778846305741715e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "grad_norm": 0.4976121485233307, + "learning_rate": 8.767998114139918e-05, + "loss": 0.0518, + "step": 2670 + }, + { + "grad_norm": 0.6053365468978882, + "learning_rate": 8.757108716314429e-05, + "loss": 0.0545, + "step": 2680 + }, + { + "grad_norm": 0.5506815910339355, + "learning_rate": 8.746178231349962e-05, + "loss": 0.0498, + "step": 2690 + }, + { + "grad_norm": 0.6330425143241882, + "learning_rate": 8.735206778780549e-05, + "loss": 0.0562, + "step": 2700 + }, + { + "grad_norm": 0.7903260588645935, + "learning_rate": 8.724194478588234e-05, + "loss": 0.0532, + "step": 2710 + }, + { + "grad_norm": 0.6343511343002319, + "learning_rate": 8.713141451201772e-05, + "loss": 0.0533, + "step": 2720 + }, + { + "grad_norm": 0.5662796497344971, + "learning_rate": 8.702047817495295e-05, + "loss": 0.0516, + "step": 2730 + }, + { + "grad_norm": 0.7495650053024292, + "learning_rate": 8.69091369878701e-05, + "loss": 0.0505, + "step": 2740 + }, + { + "grad_norm": 0.5433518290519714, + "learning_rate": 8.679739216837849e-05, + "loss": 0.0474, + "step": 2750 + }, + { + "grad_norm": 0.5403419137001038, + "learning_rate": 8.66852449385016e-05, + "loss": 0.0476, + "step": 2760 + }, + { + "grad_norm": 0.5262542963027954, + "learning_rate": 8.657269652466356e-05, + "loss": 0.0495, + "step": 2770 + }, + { + "grad_norm": 0.6162612438201904, + "learning_rate": 8.645974815767577e-05, + "loss": 0.0481, + "step": 2780 + }, + { + "grad_norm": 0.6234269142150879, + "learning_rate": 8.634640107272351e-05, + "loss": 0.0461, + "step": 2790 + }, + { + "grad_norm": 0.5556614398956299, + "learning_rate": 8.623265650935234e-05, + "loss": 0.0489, + "step": 2800 + }, + { + "grad_norm": 0.5489911437034607, + "learning_rate": 8.611851571145456e-05, + "loss": 0.0509, + "step": 2810 + }, + { + "grad_norm": 0.6572837233543396, + "learning_rate": 8.600397992725566e-05, + "loss": 0.0469, + "step": 2820 + }, + { + "grad_norm": 0.47838884592056274, + "learning_rate": 8.588905040930061e-05, + "loss": 0.0473, + "step": 2830 + }, + { + "grad_norm": 0.6445584893226624, + "learning_rate": 8.577372841444022e-05, + "loss": 0.0491, + "step": 2840 + }, + { + "grad_norm": 0.7070618271827698, + "learning_rate": 8.565801520381736e-05, + "loss": 0.0523, + "step": 2850 + }, + { + "grad_norm": 0.5014833807945251, + "learning_rate": 8.554191204285313e-05, + "loss": 0.0504, + "step": 2860 + }, + { + "grad_norm": 0.5302914381027222, + "learning_rate": 8.542542020123315e-05, + "loss": 0.0477, + "step": 2870 + }, + { + "grad_norm": 0.547792375087738, + "learning_rate": 8.530854095289347e-05, + "loss": 0.0449, + "step": 2880 + }, + { + "grad_norm": 0.6502813100814819, + "learning_rate": 8.519127557600688e-05, + "loss": 0.0478, + "step": 2890 + }, + { + "grad_norm": 0.6726429462432861, + "learning_rate": 8.507362535296871e-05, + "loss": 0.0509, + "step": 2900 + }, + { + "grad_norm": 0.5717934966087341, + "learning_rate": 8.495559157038299e-05, + "loss": 0.0646, + "step": 2910 + }, + { + "grad_norm": 0.5796350836753845, + "learning_rate": 8.483717551904823e-05, + "loss": 0.0482, + "step": 2920 + }, + { + "grad_norm": 0.541770339012146, + "learning_rate": 8.47183784939434e-05, + "loss": 0.0464, + "step": 2930 + }, + { + "grad_norm": 0.46838489174842834, + "learning_rate": 8.459920179421374e-05, + "loss": 0.0461, + "step": 2940 + }, + { + "grad_norm": 0.5526677370071411, + "learning_rate": 8.447964672315656e-05, + "loss": 0.0454, + "step": 2950 + }, + { + "grad_norm": 0.6302189826965332, + "learning_rate": 8.435971458820692e-05, + "loss": 0.0544, + "step": 2960 + }, + { + "grad_norm": 0.5775927305221558, + "learning_rate": 8.423940670092345e-05, + "loss": 0.0503, + "step": 2970 + }, + { + "grad_norm": 0.45672690868377686, + "learning_rate": 8.411872437697394e-05, + "loss": 0.0443, + "step": 2980 + }, + { + "grad_norm": 0.5356625318527222, + "learning_rate": 8.399766893612096e-05, + "loss": 0.0432, + "step": 2990 + }, + { + "grad_norm": 0.46423062682151794, + "learning_rate": 8.38762417022074e-05, + "loss": 0.0496, + "step": 3000 + }, + { + "grad_norm": 0.5346227288246155, + "learning_rate": 8.375444400314204e-05, + "loss": 0.0497, + "step": 3010 + }, + { + "grad_norm": 0.6721541285514832, + "learning_rate": 8.3632277170885e-05, + "loss": 0.0568, + "step": 3020 + }, + { + "grad_norm": 0.5497980713844299, + "learning_rate": 8.350974254143318e-05, + "loss": 0.0494, + "step": 3030 + }, + { + "grad_norm": 0.4417865574359894, + "learning_rate": 8.338684145480566e-05, + "loss": 0.0426, + "step": 3040 + }, + { + "grad_norm": 0.5426268577575684, + "learning_rate": 8.326357525502904e-05, + "loss": 0.0496, + "step": 3050 + }, + { + "grad_norm": 0.5102850794792175, + "learning_rate": 8.313994529012273e-05, + "loss": 0.0464, + "step": 3060 + }, + { + "grad_norm": 0.6222580075263977, + "learning_rate": 8.301595291208422e-05, + "loss": 0.0499, + "step": 3070 + }, + { + "grad_norm": 0.5056414604187012, + "learning_rate": 8.289159947687427e-05, + "loss": 0.0438, + "step": 3080 + }, + { + "grad_norm": 0.5930989384651184, + "learning_rate": 8.276688634440216e-05, + "loss": 0.0455, + "step": 3090 + }, + { + "grad_norm": 0.49086058139801025, + "learning_rate": 8.26418148785107e-05, + "loss": 0.0466, + "step": 3100 + }, + { + "grad_norm": 0.6659104824066162, + "learning_rate": 8.251638644696141e-05, + "loss": 0.0551, + "step": 3110 + }, + { + "grad_norm": 0.4735563099384308, + "learning_rate": 8.23906024214195e-05, + "loss": 0.044, + "step": 3120 + }, + { + "grad_norm": 0.5951749682426453, + "learning_rate": 8.226446417743897e-05, + "loss": 0.0513, + "step": 3130 + }, + { + "grad_norm": 0.4567015469074249, + "learning_rate": 8.213797309444742e-05, + "loss": 0.0491, + "step": 3140 + }, + { + "grad_norm": 0.5473612546920776, + "learning_rate": 8.201113055573105e-05, + "loss": 0.0487, + "step": 3150 + }, + { + "grad_norm": 0.5695503950119019, + "learning_rate": 8.188393794841958e-05, + "loss": 0.0445, + "step": 3160 + }, + { + "grad_norm": 0.6429879069328308, + "learning_rate": 8.175639666347094e-05, + "loss": 0.0468, + "step": 3170 + }, + { + "grad_norm": 0.532501757144928, + "learning_rate": 8.162850809565623e-05, + "loss": 0.0467, + "step": 3180 + }, + { + "grad_norm": 0.5507548451423645, + "learning_rate": 8.150027364354431e-05, + "loss": 0.0435, + "step": 3190 + }, + { + "grad_norm": 0.5140865445137024, + "learning_rate": 8.137169470948662e-05, + "loss": 0.0489, + "step": 3200 + }, + { + "grad_norm": 0.5697977542877197, + "learning_rate": 8.124277269960179e-05, + "loss": 0.0473, + "step": 3210 + }, + { + "grad_norm": 0.47149255871772766, + "learning_rate": 8.111350902376023e-05, + "loss": 0.0434, + "step": 3220 + }, + { + "grad_norm": 0.5333671569824219, + "learning_rate": 8.098390509556883e-05, + "loss": 0.0453, + "step": 3230 + }, + { + "grad_norm": 0.5352486968040466, + "learning_rate": 8.085396233235536e-05, + "loss": 0.0412, + "step": 3240 + }, + { + "grad_norm": 0.5225903391838074, + "learning_rate": 8.072368215515306e-05, + "loss": 0.0415, + "step": 3250 + }, + { + "grad_norm": 0.5469233393669128, + "learning_rate": 8.059306598868506e-05, + "loss": 0.0589, + "step": 3260 + }, + { + "grad_norm": 0.5169219970703125, + "learning_rate": 8.046211526134888e-05, + "loss": 0.0478, + "step": 3270 + }, + { + "grad_norm": 0.5857186913490295, + "learning_rate": 8.033083140520065e-05, + "loss": 0.0428, + "step": 3280 + }, + { + "grad_norm": 0.3606002628803253, + "learning_rate": 8.019921585593962e-05, + "loss": 0.045, + "step": 3290 + }, + { + "grad_norm": 0.5319432616233826, + "learning_rate": 8.006727005289232e-05, + "loss": 0.0423, + "step": 3300 + }, + { + "grad_norm": 0.37546828389167786, + "learning_rate": 7.993499543899692e-05, + "loss": 0.0467, + "step": 3310 + }, + { + "grad_norm": 0.46692955493927, + "learning_rate": 7.980239346078742e-05, + "loss": 0.0409, + "step": 3320 + }, + { + "grad_norm": 0.6998544335365295, + "learning_rate": 7.966946556837778e-05, + "loss": 0.0459, + "step": 3330 + }, + { + "grad_norm": 0.6966269612312317, + "learning_rate": 7.953621321544616e-05, + "loss": 0.0421, + "step": 3340 + }, + { + "grad_norm": 0.4158872067928314, + "learning_rate": 7.940263785921896e-05, + "loss": 0.0507, + "step": 3350 + }, + { + "grad_norm": 0.4756423532962799, + "learning_rate": 7.926874096045482e-05, + "loss": 0.0417, + "step": 3360 + }, + { + "grad_norm": 0.4818204641342163, + "learning_rate": 7.913452398342881e-05, + "loss": 0.0547, + "step": 3370 + }, + { + "grad_norm": 0.4766354560852051, + "learning_rate": 7.89999883959163e-05, + "loss": 0.0456, + "step": 3380 + }, + { + "grad_norm": 0.6075718998908997, + "learning_rate": 7.886513566917687e-05, + "loss": 0.0462, + "step": 3390 + }, + { + "grad_norm": 0.7231699824333191, + "learning_rate": 7.872996727793838e-05, + "loss": 0.0517, + "step": 3400 + }, + { + "grad_norm": 0.477040559053421, + "learning_rate": 7.859448470038069e-05, + "loss": 0.0432, + "step": 3410 + }, + { + "grad_norm": 0.4753010869026184, + "learning_rate": 7.845868941811956e-05, + "loss": 0.0429, + "step": 3420 + }, + { + "grad_norm": 0.4999731481075287, + "learning_rate": 7.832258291619043e-05, + "loss": 0.0413, + "step": 3430 + }, + { + "grad_norm": 0.6600528359413147, + "learning_rate": 7.81861666830322e-05, + "loss": 0.0442, + "step": 3440 + }, + { + "grad_norm": 0.49738991260528564, + "learning_rate": 7.804944221047097e-05, + "loss": 0.0399, + "step": 3450 + }, + { + "grad_norm": 0.5758061408996582, + "learning_rate": 7.791241099370364e-05, + "loss": 0.0499, + "step": 3460 + }, + { + "grad_norm": 0.48965656757354736, + "learning_rate": 7.777507453128163e-05, + "loss": 0.0383, + "step": 3470 + }, + { + "grad_norm": 0.5145290493965149, + "learning_rate": 7.763743432509451e-05, + "loss": 0.0478, + "step": 3480 + }, + { + "grad_norm": 0.5944988131523132, + "learning_rate": 7.749949188035353e-05, + "loss": 0.0421, + "step": 3490 + }, + { + "grad_norm": 0.5881758332252502, + "learning_rate": 7.736124870557516e-05, + "loss": 0.0504, + "step": 3500 + }, + { + "grad_norm": 0.4876430034637451, + "learning_rate": 7.722270631256459e-05, + "loss": 0.0425, + "step": 3510 + }, + { + "grad_norm": 0.5497170686721802, + "learning_rate": 7.708386621639925e-05, + "loss": 0.0406, + "step": 3520 + }, + { + "grad_norm": 0.5289949178695679, + "learning_rate": 7.694472993541219e-05, + "loss": 0.0432, + "step": 3530 + }, + { + "grad_norm": 0.5657542943954468, + "learning_rate": 7.680529899117547e-05, + "loss": 0.048, + "step": 3540 + }, + { + "grad_norm": 0.6215736865997314, + "learning_rate": 7.666557490848358e-05, + "loss": 0.0435, + "step": 3550 + }, + { + "grad_norm": 0.5217798948287964, + "learning_rate": 7.65255592153367e-05, + "loss": 0.0439, + "step": 3560 + }, + { + "grad_norm": 0.5852975249290466, + "learning_rate": 7.638525344292402e-05, + "loss": 0.0433, + "step": 3570 + }, + { + "grad_norm": 0.5029234886169434, + "learning_rate": 7.624465912560697e-05, + "loss": 0.042, + "step": 3580 + }, + { + "grad_norm": 0.46533602476119995, + "learning_rate": 7.610377780090249e-05, + "loss": 0.0423, + "step": 3590 + }, + { + "grad_norm": 0.5186182856559753, + "learning_rate": 7.596261100946618e-05, + "loss": 0.0464, + "step": 3600 + }, + { + "grad_norm": 0.6336029171943665, + "learning_rate": 7.582116029507542e-05, + "loss": 0.046, + "step": 3610 + }, + { + "grad_norm": 0.5276234745979309, + "learning_rate": 7.56794272046126e-05, + "loss": 0.0463, + "step": 3620 + }, + { + "grad_norm": 0.47521013021469116, + "learning_rate": 7.55374132880481e-05, + "loss": 0.0482, + "step": 3630 + }, + { + "grad_norm": 0.4061400592327118, + "learning_rate": 7.539512009842333e-05, + "loss": 0.0486, + "step": 3640 + }, + { + "grad_norm": 0.5201972126960754, + "learning_rate": 7.525254919183382e-05, + "loss": 0.0459, + "step": 3650 + }, + { + "grad_norm": 0.4960852563381195, + "learning_rate": 7.510970212741215e-05, + "loss": 0.0407, + "step": 3660 + }, + { + "grad_norm": 0.6205229163169861, + "learning_rate": 7.496658046731096e-05, + "loss": 0.0438, + "step": 3670 + }, + { + "grad_norm": 0.5703109502792358, + "learning_rate": 7.482318577668578e-05, + "loss": 0.046, + "step": 3680 + }, + { + "grad_norm": 0.5233997702598572, + "learning_rate": 7.467951962367796e-05, + "loss": 0.0474, + "step": 3690 + }, + { + "grad_norm": 0.5744036436080933, + "learning_rate": 7.453558357939755e-05, + "loss": 0.0455, + "step": 3700 + }, + { + "grad_norm": 0.5415105223655701, + "learning_rate": 7.439137921790606e-05, + "loss": 0.0495, + "step": 3710 + }, + { + "grad_norm": 0.45560383796691895, + "learning_rate": 7.42469081161993e-05, + "loss": 0.0465, + "step": 3720 + }, + { + "grad_norm": 0.4683740437030792, + "learning_rate": 7.410217185419006e-05, + "loss": 0.0432, + "step": 3730 + }, + { + "grad_norm": 0.4923056364059448, + "learning_rate": 7.395717201469095e-05, + "loss": 0.0492, + "step": 3740 + }, + { + "grad_norm": 0.5593320727348328, + "learning_rate": 7.381191018339696e-05, + "loss": 0.0444, + "step": 3750 + }, + { + "grad_norm": 0.4336199164390564, + "learning_rate": 7.36663879488682e-05, + "loss": 0.0463, + "step": 3760 + }, + { + "grad_norm": 0.4620218575000763, + "learning_rate": 7.352060690251254e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "grad_norm": 0.592485785484314, + "learning_rate": 7.337456863856811e-05, + "loss": 0.0422, + "step": 3780 + }, + { + "grad_norm": 0.6306523680686951, + "learning_rate": 7.3228274754086e-05, + "loss": 0.0423, + "step": 3790 + }, + { + "grad_norm": 0.5346505045890808, + "learning_rate": 7.308172684891267e-05, + "loss": 0.0417, + "step": 3800 + }, + { + "grad_norm": 0.4896720051765442, + "learning_rate": 7.293492652567255e-05, + "loss": 0.0382, + "step": 3810 + }, + { + "grad_norm": 0.4991335868835449, + "learning_rate": 7.278787538975043e-05, + "loss": 0.0439, + "step": 3820 + }, + { + "grad_norm": 0.4827921390533447, + "learning_rate": 7.2640575049274e-05, + "loss": 0.0435, + "step": 3830 + }, + { + "grad_norm": 0.4323563277721405, + "learning_rate": 7.249302711509616e-05, + "loss": 0.0431, + "step": 3840 + }, + { + "grad_norm": 0.5075383186340332, + "learning_rate": 7.23452332007775e-05, + "loss": 0.0411, + "step": 3850 + }, + { + "grad_norm": 0.46734702587127686, + "learning_rate": 7.219719492256858e-05, + "loss": 0.0446, + "step": 3860 + }, + { + "grad_norm": 0.6235191226005554, + "learning_rate": 7.20489138993923e-05, + "loss": 0.0432, + "step": 3870 + }, + { + "grad_norm": 0.435124009847641, + "learning_rate": 7.190039175282614e-05, + "loss": 0.0429, + "step": 3880 + }, + { + "grad_norm": 0.39774423837661743, + "learning_rate": 7.175163010708455e-05, + "loss": 0.0377, + "step": 3890 + }, + { + "grad_norm": 0.5165033340454102, + "learning_rate": 7.1602630589001e-05, + "loss": 0.0409, + "step": 3900 + }, + { + "grad_norm": 0.43759685754776, + "learning_rate": 7.14533948280104e-05, + "loss": 0.041, + "step": 3910 + }, + { + "grad_norm": 0.5358320474624634, + "learning_rate": 7.130392445613109e-05, + "loss": 0.0395, + "step": 3920 + }, + { + "grad_norm": 0.49284276366233826, + "learning_rate": 7.115422110794711e-05, + "loss": 0.0382, + "step": 3930 + }, + { + "grad_norm": 0.6859235167503357, + "learning_rate": 7.100428642059033e-05, + "loss": 0.04, + "step": 3940 + }, + { + "grad_norm": 0.543547511100769, + "learning_rate": 7.08541220337224e-05, + "loss": 0.0449, + "step": 3950 + }, + { + "grad_norm": 0.42491382360458374, + "learning_rate": 7.070372958951706e-05, + "loss": 0.0399, + "step": 3960 + }, + { + "grad_norm": 0.4901983439922333, + "learning_rate": 7.055311073264194e-05, + "loss": 0.0435, + "step": 3970 + }, + { + "grad_norm": 0.5592600703239441, + "learning_rate": 7.040226711024077e-05, + "loss": 0.0376, + "step": 3980 + }, + { + "grad_norm": 0.5757395625114441, + "learning_rate": 7.02512003719152e-05, + "loss": 0.0441, + "step": 3990 + }, + { + "grad_norm": 0.6468963027000427, + "learning_rate": 7.00999121697069e-05, + "loss": 0.0448, + "step": 4000 + }, + { + "grad_norm": 0.5414806008338928, + "learning_rate": 6.99484041580794e-05, + "loss": 0.0428, + "step": 4010 + }, + { + "grad_norm": 0.46762150526046753, + "learning_rate": 6.979667799390004e-05, + "loss": 0.0388, + "step": 4020 + }, + { + "grad_norm": 0.53855299949646, + "learning_rate": 6.964473533642185e-05, + "loss": 0.044, + "step": 4030 + }, + { + "grad_norm": 0.48193594813346863, + "learning_rate": 6.949257784726539e-05, + "loss": 0.0408, + "step": 4040 + }, + { + "grad_norm": 0.5546448230743408, + "learning_rate": 6.934020719040056e-05, + "loss": 0.039, + "step": 4050 + }, + { + "grad_norm": 0.636574387550354, + "learning_rate": 6.918762503212848e-05, + "loss": 0.0453, + "step": 4060 + }, + { + "grad_norm": 0.5646332502365112, + "learning_rate": 6.903483304106319e-05, + "loss": 0.0466, + "step": 4070 + }, + { + "grad_norm": 0.5054908990859985, + "learning_rate": 6.888183288811341e-05, + "loss": 0.0425, + "step": 4080 + }, + { + "grad_norm": 0.5238390564918518, + "learning_rate": 6.87286262464643e-05, + "loss": 0.0411, + "step": 4090 + }, + { + "grad_norm": 0.6083115339279175, + "learning_rate": 6.857521479155915e-05, + "loss": 0.0466, + "step": 4100 + }, + { + "grad_norm": 0.5463563203811646, + "learning_rate": 6.842160020108104e-05, + "loss": 0.0409, + "step": 4110 + }, + { + "grad_norm": 0.46191924810409546, + "learning_rate": 6.826778415493455e-05, + "loss": 0.0401, + "step": 4120 + }, + { + "grad_norm": 0.47926998138427734, + "learning_rate": 6.811376833522729e-05, + "loss": 0.0385, + "step": 4130 + }, + { + "grad_norm": 0.5727494359016418, + "learning_rate": 6.795955442625159e-05, + "loss": 0.0479, + "step": 4140 + }, + { + "grad_norm": 0.4484681487083435, + "learning_rate": 6.780514411446608e-05, + "loss": 0.0415, + "step": 4150 + }, + { + "grad_norm": 0.483377605676651, + "learning_rate": 6.765053908847716e-05, + "loss": 0.04, + "step": 4160 + }, + { + "grad_norm": 0.48677024245262146, + "learning_rate": 6.749574103902064e-05, + "loss": 0.044, + "step": 4170 + }, + { + "grad_norm": 0.5038551092147827, + "learning_rate": 6.734075165894317e-05, + "loss": 0.0426, + "step": 4180 + }, + { + "grad_norm": 0.5084865689277649, + "learning_rate": 6.71855726431838e-05, + "loss": 0.0419, + "step": 4190 + }, + { + "grad_norm": 0.5698152780532837, + "learning_rate": 6.703020568875538e-05, + "loss": 0.0365, + "step": 4200 + }, + { + "grad_norm": 0.5380019545555115, + "learning_rate": 6.687465249472603e-05, + "loss": 0.0433, + "step": 4210 + }, + { + "grad_norm": 0.6034384965896606, + "learning_rate": 6.671891476220055e-05, + "loss": 0.0376, + "step": 4220 + }, + { + "grad_norm": 0.43863195180892944, + "learning_rate": 6.656299419430183e-05, + "loss": 0.0384, + "step": 4230 + }, + { + "grad_norm": 0.3712943196296692, + "learning_rate": 6.640689249615223e-05, + "loss": 0.0398, + "step": 4240 + }, + { + "grad_norm": 0.5271754860877991, + "learning_rate": 6.625061137485491e-05, + "loss": 0.0395, + "step": 4250 + }, + { + "grad_norm": 0.3926122784614563, + "learning_rate": 6.609415253947517e-05, + "loss": 0.0405, + "step": 4260 + }, + { + "grad_norm": 0.3551841080188751, + "learning_rate": 6.593751770102178e-05, + "loss": 0.0394, + "step": 4270 + }, + { + "grad_norm": 0.514536440372467, + "learning_rate": 6.578070857242823e-05, + "loss": 0.0398, + "step": 4280 + }, + { + "grad_norm": 0.47524216771125793, + "learning_rate": 6.562372686853402e-05, + "loss": 0.0381, + "step": 4290 + }, + { + "grad_norm": 0.5025336742401123, + "learning_rate": 6.546657430606593e-05, + "loss": 0.04, + "step": 4300 + }, + { + "grad_norm": 0.4409259855747223, + "learning_rate": 6.530925260361918e-05, + "loss": 0.0372, + "step": 4310 + }, + { + "grad_norm": 0.4905289113521576, + "learning_rate": 6.515176348163871e-05, + "loss": 0.0418, + "step": 4320 + }, + { + "grad_norm": 0.3949728012084961, + "learning_rate": 6.499410866240032e-05, + "loss": 0.0391, + "step": 4330 + }, + { + "grad_norm": 0.4823022186756134, + "learning_rate": 6.48362898699919e-05, + "loss": 0.0359, + "step": 4340 + }, + { + "grad_norm": 0.48058265447616577, + "learning_rate": 6.467830883029443e-05, + "loss": 0.0382, + "step": 4350 + }, + { + "grad_norm": 0.5487723350524902, + "learning_rate": 6.452016727096326e-05, + "loss": 0.0366, + "step": 4360 + }, + { + "grad_norm": 0.48702216148376465, + "learning_rate": 6.436186692140916e-05, + "loss": 0.0376, + "step": 4370 + }, + { + "grad_norm": 0.4363824129104614, + "learning_rate": 6.420340951277938e-05, + "loss": 0.0386, + "step": 4380 + }, + { + "grad_norm": 0.5320507287979126, + "learning_rate": 6.404479677793874e-05, + "loss": 0.0434, + "step": 4390 + }, + { + "grad_norm": 0.5952131152153015, + "learning_rate": 6.388603045145075e-05, + "loss": 0.044, + "step": 4400 + }, + { + "grad_norm": 0.4177091717720032, + "learning_rate": 6.372711226955843e-05, + "loss": 0.0373, + "step": 4410 + }, + { + "grad_norm": 0.5674641728401184, + "learning_rate": 6.356804397016564e-05, + "loss": 0.0437, + "step": 4420 + }, + { + "grad_norm": 0.472881555557251, + "learning_rate": 6.340882729281779e-05, + "loss": 0.0416, + "step": 4430 + }, + { + "grad_norm": 0.5012088418006897, + "learning_rate": 6.324946397868294e-05, + "loss": 0.0457, + "step": 4440 + }, + { + "grad_norm": 0.49250346422195435, + "learning_rate": 6.308995577053276e-05, + "loss": 0.0358, + "step": 4450 + }, + { + "grad_norm": 0.442531943321228, + "learning_rate": 6.293030441272347e-05, + "loss": 0.0391, + "step": 4460 + }, + { + "grad_norm": 0.45584627985954285, + "learning_rate": 6.277051165117677e-05, + "loss": 0.036, + "step": 4470 + }, + { + "grad_norm": 0.46474725008010864, + "learning_rate": 6.261057923336064e-05, + "loss": 0.0353, + "step": 4480 + }, + { + "grad_norm": 0.49693354964256287, + "learning_rate": 6.245050890827042e-05, + "loss": 0.0404, + "step": 4490 + }, + { + "grad_norm": 0.4872795045375824, + "learning_rate": 6.229030242640952e-05, + "loss": 0.0408, + "step": 4500 + }, + { + "grad_norm": 0.6823276877403259, + "learning_rate": 6.212996153977037e-05, + "loss": 0.0458, + "step": 4510 + }, + { + "grad_norm": 0.4600830376148224, + "learning_rate": 6.196948800181523e-05, + "loss": 0.0379, + "step": 4520 + }, + { + "grad_norm": 0.5189037322998047, + "learning_rate": 6.180888356745695e-05, + "loss": 0.0391, + "step": 4530 + }, + { + "grad_norm": 0.40244150161743164, + "learning_rate": 6.164814999303995e-05, + "loss": 0.04, + "step": 4540 + }, + { + "grad_norm": 0.4358130991458893, + "learning_rate": 6.148728903632081e-05, + "loss": 0.0368, + "step": 4550 + }, + { + "grad_norm": 0.5342580080032349, + "learning_rate": 6.132630245644921e-05, + "loss": 0.0344, + "step": 4560 + }, + { + "grad_norm": 0.5356363654136658, + "learning_rate": 6.116519201394857e-05, + "loss": 0.0376, + "step": 4570 + }, + { + "grad_norm": 0.4631805419921875, + "learning_rate": 6.10039594706969e-05, + "loss": 0.0376, + "step": 4580 + }, + { + "grad_norm": 0.43799665570259094, + "learning_rate": 6.084260658990744e-05, + "loss": 0.0377, + "step": 4590 + }, + { + "grad_norm": 0.4513268768787384, + "learning_rate": 6.068113513610943e-05, + "loss": 0.0364, + "step": 4600 + }, + { + "grad_norm": 0.5678306221961975, + "learning_rate": 6.0519546875128876e-05, + "loss": 0.0373, + "step": 4610 + }, + { + "grad_norm": 0.476609468460083, + "learning_rate": 6.035784357406906e-05, + "loss": 0.0403, + "step": 4620 + }, + { + "grad_norm": 0.4625353217124939, + "learning_rate": 6.01960270012914e-05, + "loss": 0.0376, + "step": 4630 + }, + { + "grad_norm": 0.4942452311515808, + "learning_rate": 6.003409892639599e-05, + "loss": 0.0375, + "step": 4640 + }, + { + "grad_norm": 0.46605247259140015, + "learning_rate": 5.9872061120202336e-05, + "loss": 0.0393, + "step": 4650 + }, + { + "grad_norm": 0.531728208065033, + "learning_rate": 5.9709915354729914e-05, + "loss": 0.0473, + "step": 4660 + }, + { + "grad_norm": 0.4280456602573395, + "learning_rate": 5.9547663403178824e-05, + "loss": 0.0378, + "step": 4670 + }, + { + "grad_norm": 0.42963069677352905, + "learning_rate": 5.9385307039910445e-05, + "loss": 0.0321, + "step": 4680 + }, + { + "grad_norm": 0.4135236144065857, + "learning_rate": 5.922284804042792e-05, + "loss": 0.0417, + "step": 4690 + }, + { + "grad_norm": 0.5570497512817383, + "learning_rate": 5.906028818135687e-05, + "loss": 0.0439, + "step": 4700 + }, + { + "grad_norm": 0.41998130083084106, + "learning_rate": 5.889762924042585e-05, + "loss": 0.0365, + "step": 4710 + }, + { + "grad_norm": 0.543883740901947, + "learning_rate": 5.873487299644699e-05, + "loss": 0.0352, + "step": 4720 + }, + { + "grad_norm": 0.467123419046402, + "learning_rate": 5.857202122929649e-05, + "loss": 0.0376, + "step": 4730 + }, + { + "grad_norm": 0.4760608971118927, + "learning_rate": 5.840907571989518e-05, + "loss": 0.034, + "step": 4740 + }, + { + "grad_norm": 0.501928985118866, + "learning_rate": 5.824603825018904e-05, + "loss": 0.0379, + "step": 4750 + }, + { + "grad_norm": 0.5355032682418823, + "learning_rate": 5.808291060312975e-05, + "loss": 0.0364, + "step": 4760 + }, + { + "grad_norm": 0.5742557048797607, + "learning_rate": 5.7919694562655083e-05, + "loss": 0.0381, + "step": 4770 + }, + { + "grad_norm": 0.5819830894470215, + "learning_rate": 5.775639191366954e-05, + "loss": 0.0382, + "step": 4780 + }, + { + "grad_norm": 0.3822622299194336, + "learning_rate": 5.75930044420247e-05, + "loss": 0.0375, + "step": 4790 + }, + { + "grad_norm": 0.4396764039993286, + "learning_rate": 5.74295339344998e-05, + "loss": 0.0398, + "step": 4800 + }, + { + "grad_norm": 0.4816092848777771, + "learning_rate": 5.726598217878211e-05, + "loss": 0.0389, + "step": 4810 + }, + { + "grad_norm": 0.3898105025291443, + "learning_rate": 5.71023509634474e-05, + "loss": 0.0343, + "step": 4820 + }, + { + "grad_norm": 0.44770386815071106, + "learning_rate": 5.693864207794049e-05, + "loss": 0.0379, + "step": 4830 + }, + { + "grad_norm": 0.4150899648666382, + "learning_rate": 5.677485731255545e-05, + "loss": 0.0433, + "step": 4840 + }, + { + "grad_norm": 0.4779394865036011, + "learning_rate": 5.6610998458416296e-05, + "loss": 0.0374, + "step": 4850 + }, + { + "grad_norm": 0.4998456537723541, + "learning_rate": 5.644706730745716e-05, + "loss": 0.0383, + "step": 4860 + }, + { + "grad_norm": 0.3862738311290741, + "learning_rate": 5.628306565240287e-05, + "loss": 0.0358, + "step": 4870 + }, + { + "grad_norm": 0.43701767921447754, + "learning_rate": 5.611899528674923e-05, + "loss": 0.0385, + "step": 4880 + }, + { + "grad_norm": 0.5248092412948608, + "learning_rate": 5.595485800474349e-05, + "loss": 0.0403, + "step": 4890 + }, + { + "grad_norm": 0.49775761365890503, + "learning_rate": 5.579065560136467e-05, + "loss": 0.0387, + "step": 4900 + }, + { + "grad_norm": 0.38317522406578064, + "learning_rate": 5.562638987230392e-05, + "loss": 0.0336, + "step": 4910 + }, + { + "grad_norm": 0.48581790924072266, + "learning_rate": 5.546206261394498e-05, + "loss": 0.0325, + "step": 4920 + }, + { + "grad_norm": 0.4865744113922119, + "learning_rate": 5.529767562334437e-05, + "loss": 0.036, + "step": 4930 + }, + { + "grad_norm": 0.5772055983543396, + "learning_rate": 5.5133230698211926e-05, + "loss": 0.0326, + "step": 4940 + }, + { + "grad_norm": 0.3783220648765564, + "learning_rate": 5.496872963689096e-05, + "loss": 0.036, + "step": 4950 + }, + { + "grad_norm": 0.4435168504714966, + "learning_rate": 5.4804174238338756e-05, + "loss": 0.0345, + "step": 4960 + }, + { + "grad_norm": 0.4597376585006714, + "learning_rate": 5.463956630210678e-05, + "loss": 0.0427, + "step": 4970 + }, + { + "grad_norm": 0.41899651288986206, + "learning_rate": 5.4474907628321046e-05, + "loss": 0.0401, + "step": 4980 + }, + { + "grad_norm": 0.5019888877868652, + "learning_rate": 5.431020001766244e-05, + "loss": 0.0342, + "step": 4990 + }, + { + "grad_norm": 0.490264892578125, + "learning_rate": 5.4145445271346986e-05, + "loss": 0.0347, + "step": 5000 + }, + { + "grad_norm": 0.4609229564666748, + "learning_rate": 5.398064519110622e-05, + "loss": 0.0351, + "step": 5010 + }, + { + "grad_norm": 0.4674476683139801, + "learning_rate": 5.3815801579167394e-05, + "loss": 0.0362, + "step": 5020 + }, + { + "grad_norm": 0.5679224729537964, + "learning_rate": 5.365091623823382e-05, + "loss": 0.0354, + "step": 5030 + }, + { + "grad_norm": 0.4846169054508209, + "learning_rate": 5.348599097146521e-05, + "loss": 0.036, + "step": 5040 + }, + { + "grad_norm": 0.45144814252853394, + "learning_rate": 5.3321027582457836e-05, + "loss": 0.0377, + "step": 5050 + }, + { + "grad_norm": 0.4690643846988678, + "learning_rate": 5.315602787522491e-05, + "loss": 0.0381, + "step": 5060 + }, + { + "grad_norm": 0.4994000494480133, + "learning_rate": 5.299099365417678e-05, + "loss": 0.0333, + "step": 5070 + }, + { + "grad_norm": 0.464571475982666, + "learning_rate": 5.2825926724101236e-05, + "loss": 0.0357, + "step": 5080 + }, + { + "grad_norm": 0.5771486163139343, + "learning_rate": 5.26608288901438e-05, + "loss": 0.0347, + "step": 5090 + }, + { + "grad_norm": 0.5044403672218323, + "learning_rate": 5.24957019577879e-05, + "loss": 0.0372, + "step": 5100 + }, + { + "grad_norm": 0.4808100461959839, + "learning_rate": 5.2330547732835266e-05, + "loss": 0.0337, + "step": 5110 + }, + { + "grad_norm": 0.5282689929008484, + "learning_rate": 5.2165368021385996e-05, + "loss": 0.0452, + "step": 5120 + }, + { + "grad_norm": 0.36428406834602356, + "learning_rate": 5.200016462981897e-05, + "loss": 0.0322, + "step": 5130 + }, + { + "grad_norm": 0.44125813245773315, + "learning_rate": 5.1834939364772015e-05, + "loss": 0.0325, + "step": 5140 + }, + { + "grad_norm": 0.41744136810302734, + "learning_rate": 5.166969403312214e-05, + "loss": 0.0319, + "step": 5150 + }, + { + "grad_norm": 0.42846477031707764, + "learning_rate": 5.1504430441965844e-05, + "loss": 0.032, + "step": 5160 + }, + { + "grad_norm": 0.4896695911884308, + "learning_rate": 5.133915039859923e-05, + "loss": 0.0296, + "step": 5170 + }, + { + "grad_norm": 0.511642336845398, + "learning_rate": 5.1173855710498444e-05, + "loss": 0.0333, + "step": 5180 + }, + { + "grad_norm": 0.45660096406936646, + "learning_rate": 5.100854818529967e-05, + "loss": 0.0337, + "step": 5190 + }, + { + "grad_norm": 0.4637276828289032, + "learning_rate": 5.084322963077951e-05, + "loss": 0.0396, + "step": 5200 + }, + { + "grad_norm": 0.3938063979148865, + "learning_rate": 5.067790185483522e-05, + "loss": 0.0381, + "step": 5210 + }, + { + "grad_norm": 0.40539443492889404, + "learning_rate": 5.0512566665464844e-05, + "loss": 0.0348, + "step": 5220 + }, + { + "grad_norm": 0.38393694162368774, + "learning_rate": 5.034722587074755e-05, + "loss": 0.0339, + "step": 5230 + }, + { + "grad_norm": 0.4826929569244385, + "learning_rate": 5.018188127882375e-05, + "loss": 0.0342, + "step": 5240 + }, + { + "grad_norm": 0.45363306999206543, + "learning_rate": 5.0016534697875417e-05, + "loss": 0.0302, + "step": 5250 + }, + { + "grad_norm": 0.40788865089416504, + "learning_rate": 4.9851187936106294e-05, + "loss": 0.0372, + "step": 5260 + }, + { + "grad_norm": 0.4535907208919525, + "learning_rate": 4.968584280172206e-05, + "loss": 0.0349, + "step": 5270 + }, + { + "grad_norm": 0.4888495206832886, + "learning_rate": 4.95205011029106e-05, + "loss": 0.034, + "step": 5280 + }, + { + "grad_norm": 0.432629257440567, + "learning_rate": 4.935516464782227e-05, + "loss": 0.0341, + "step": 5290 + }, + { + "grad_norm": 0.5273644924163818, + "learning_rate": 4.918983524455003e-05, + "loss": 0.0341, + "step": 5300 + }, + { + "grad_norm": 0.5193825364112854, + "learning_rate": 4.9024514701109766e-05, + "loss": 0.0312, + "step": 5310 + }, + { + "grad_norm": 0.383564293384552, + "learning_rate": 4.885920482542043e-05, + "loss": 0.0341, + "step": 5320 + }, + { + "grad_norm": 0.3665809631347656, + "learning_rate": 4.869390742528438e-05, + "loss": 0.0298, + "step": 5330 + }, + { + "grad_norm": 0.4296775758266449, + "learning_rate": 4.852862430836744e-05, + "loss": 0.036, + "step": 5340 + }, + { + "grad_norm": 0.39151832461357117, + "learning_rate": 4.836335728217933e-05, + "loss": 0.0309, + "step": 5350 + }, + { + "grad_norm": 0.5093752145767212, + "learning_rate": 4.819810815405379e-05, + "loss": 0.0349, + "step": 5360 + }, + { + "grad_norm": 0.6182861924171448, + "learning_rate": 4.803287873112877e-05, + "loss": 0.0372, + "step": 5370 + }, + { + "grad_norm": 0.5121681094169617, + "learning_rate": 4.786767082032681e-05, + "loss": 0.0351, + "step": 5380 + }, + { + "grad_norm": 0.3878619968891144, + "learning_rate": 4.77024862283351e-05, + "loss": 0.0334, + "step": 5390 + }, + { + "grad_norm": 0.49869996309280396, + "learning_rate": 4.753732676158593e-05, + "loss": 0.0339, + "step": 5400 + }, + { + "grad_norm": 0.5226000547409058, + "learning_rate": 4.737219422623672e-05, + "loss": 0.0333, + "step": 5410 + }, + { + "grad_norm": 0.48753395676612854, + "learning_rate": 4.720709042815044e-05, + "loss": 0.0312, + "step": 5420 + }, + { + "grad_norm": 0.46787673234939575, + "learning_rate": 4.704201717287578e-05, + "loss": 0.0327, + "step": 5430 + }, + { + "grad_norm": 0.5108851194381714, + "learning_rate": 4.6876976265627404e-05, + "loss": 0.0326, + "step": 5440 + }, + { + "grad_norm": 0.5048650503158569, + "learning_rate": 4.671196951126626e-05, + "loss": 0.0376, + "step": 5450 + }, + { + "grad_norm": 0.3431260883808136, + "learning_rate": 4.654699871427971e-05, + "loss": 0.0308, + "step": 5460 + }, + { + "grad_norm": 0.4449670910835266, + "learning_rate": 4.6382065678762034e-05, + "loss": 0.0352, + "step": 5470 + }, + { + "grad_norm": 0.37903866171836853, + "learning_rate": 4.6217172208394424e-05, + "loss": 0.0322, + "step": 5480 + }, + { + "grad_norm": 0.38446661829948425, + "learning_rate": 4.605232010642549e-05, + "loss": 0.0286, + "step": 5490 + }, + { + "grad_norm": 0.46681928634643555, + "learning_rate": 4.588751117565142e-05, + "loss": 0.0324, + "step": 5500 + }, + { + "grad_norm": 0.3996520936489105, + "learning_rate": 4.5722747218396214e-05, + "loss": 0.0353, + "step": 5510 + }, + { + "grad_norm": 0.3497842848300934, + "learning_rate": 4.5558030036492194e-05, + "loss": 0.0328, + "step": 5520 + }, + { + "grad_norm": 0.4016847014427185, + "learning_rate": 4.539336143125999e-05, + "loss": 0.0381, + "step": 5530 + }, + { + "grad_norm": 0.4591330587863922, + "learning_rate": 4.522874320348916e-05, + "loss": 0.032, + "step": 5540 + }, + { + "grad_norm": 0.5982142686843872, + "learning_rate": 4.506417715341821e-05, + "loss": 0.0323, + "step": 5550 + }, + { + "grad_norm": 0.5174066424369812, + "learning_rate": 4.489966508071511e-05, + "loss": 0.0316, + "step": 5560 + }, + { + "grad_norm": 0.4469364285469055, + "learning_rate": 4.4735208784457575e-05, + "loss": 0.0288, + "step": 5570 + }, + { + "grad_norm": 0.5118666291236877, + "learning_rate": 4.457081006311325e-05, + "loss": 0.0341, + "step": 5580 + }, + { + "grad_norm": 0.495812326669693, + "learning_rate": 4.440647071452027e-05, + "loss": 0.0346, + "step": 5590 + }, + { + "grad_norm": 0.38718587160110474, + "learning_rate": 4.424219253586737e-05, + "loss": 0.0344, + "step": 5600 + }, + { + "grad_norm": 0.451130211353302, + "learning_rate": 4.407797732367443e-05, + "loss": 0.0337, + "step": 5610 + }, + { + "grad_norm": 0.43241921067237854, + "learning_rate": 4.391382687377268e-05, + "loss": 0.0328, + "step": 5620 + }, + { + "grad_norm": 0.5730195641517639, + "learning_rate": 4.374974298128512e-05, + "loss": 0.0301, + "step": 5630 + }, + { + "grad_norm": 0.4620615541934967, + "learning_rate": 4.358572744060699e-05, + "loss": 0.0398, + "step": 5640 + }, + { + "grad_norm": 0.5538221597671509, + "learning_rate": 4.342178204538588e-05, + "loss": 0.0306, + "step": 5650 + }, + { + "grad_norm": 0.5564296841621399, + "learning_rate": 4.325790858850241e-05, + "loss": 0.0328, + "step": 5660 + }, + { + "grad_norm": 0.4357956647872925, + "learning_rate": 4.309410886205043e-05, + "loss": 0.0321, + "step": 5670 + }, + { + "grad_norm": 0.30211740732192993, + "learning_rate": 4.293038465731752e-05, + "loss": 0.0325, + "step": 5680 + }, + { + "grad_norm": 0.5357030630111694, + "learning_rate": 4.276673776476533e-05, + "loss": 0.0329, + "step": 5690 + }, + { + "grad_norm": 0.4498059153556824, + "learning_rate": 4.260316997401007e-05, + "loss": 0.0287, + "step": 5700 + }, + { + "grad_norm": 0.4459388852119446, + "learning_rate": 4.243968307380293e-05, + "loss": 0.0329, + "step": 5710 + }, + { + "grad_norm": 0.40281790494918823, + "learning_rate": 4.22762788520104e-05, + "loss": 0.033, + "step": 5720 + }, + { + "grad_norm": 0.39583978056907654, + "learning_rate": 4.211295909559491e-05, + "loss": 0.0342, + "step": 5730 + }, + { + "grad_norm": 0.35422611236572266, + "learning_rate": 4.194972559059511e-05, + "loss": 0.0305, + "step": 5740 + }, + { + "grad_norm": 0.360821008682251, + "learning_rate": 4.178658012210651e-05, + "loss": 0.0282, + "step": 5750 + }, + { + "grad_norm": 0.4884454905986786, + "learning_rate": 4.162352447426177e-05, + "loss": 0.0293, + "step": 5760 + }, + { + "grad_norm": 0.4510625898838043, + "learning_rate": 4.146056043021135e-05, + "loss": 0.0314, + "step": 5770 + }, + { + "grad_norm": 0.38459086418151855, + "learning_rate": 4.1297689772103944e-05, + "loss": 0.0339, + "step": 5780 + }, + { + "grad_norm": 0.4779391884803772, + "learning_rate": 4.113491428106694e-05, + "loss": 0.0325, + "step": 5790 + }, + { + "grad_norm": 0.46030163764953613, + "learning_rate": 4.0972235737187055e-05, + "loss": 0.0289, + "step": 5800 + }, + { + "grad_norm": 0.47185903787612915, + "learning_rate": 4.080965591949076e-05, + "loss": 0.0264, + "step": 5810 + }, + { + "grad_norm": 0.5378159880638123, + "learning_rate": 4.0647176605924924e-05, + "loss": 0.0303, + "step": 5820 + }, + { + "grad_norm": 0.472313791513443, + "learning_rate": 4.0484799573337255e-05, + "loss": 0.034, + "step": 5830 + }, + { + "grad_norm": 0.40909892320632935, + "learning_rate": 4.032252659745699e-05, + "loss": 0.0296, + "step": 5840 + }, + { + "grad_norm": 0.46723711490631104, + "learning_rate": 4.016035945287539e-05, + "loss": 0.037, + "step": 5850 + }, + { + "grad_norm": 0.4663132131099701, + "learning_rate": 3.999829991302635e-05, + "loss": 0.0358, + "step": 5860 + }, + { + "grad_norm": 0.3933708071708679, + "learning_rate": 3.983634975016707e-05, + "loss": 0.0306, + "step": 5870 + }, + { + "grad_norm": 0.44689297676086426, + "learning_rate": 3.967451073535854e-05, + "loss": 0.04, + "step": 5880 + }, + { + "grad_norm": 0.507107138633728, + "learning_rate": 3.951278463844633e-05, + "loss": 0.0265, + "step": 5890 + }, + { + "grad_norm": 0.40751364827156067, + "learning_rate": 3.935117322804111e-05, + "loss": 0.0329, + "step": 5900 + }, + { + "grad_norm": 0.3654705584049225, + "learning_rate": 3.918967827149938e-05, + "loss": 0.03, + "step": 5910 + }, + { + "grad_norm": 0.38887983560562134, + "learning_rate": 3.9028301534904094e-05, + "loss": 0.0289, + "step": 5920 + }, + { + "grad_norm": 0.4749573767185211, + "learning_rate": 3.88670447830454e-05, + "loss": 0.0275, + "step": 5930 + }, + { + "grad_norm": 0.5203545093536377, + "learning_rate": 3.870590977940132e-05, + "loss": 0.0289, + "step": 5940 + }, + { + "grad_norm": 0.5058891177177429, + "learning_rate": 3.8544898286118404e-05, + "loss": 0.0263, + "step": 5950 + }, + { + "grad_norm": 0.5234279632568359, + "learning_rate": 3.838401206399257e-05, + "loss": 0.0271, + "step": 5960 + }, + { + "grad_norm": 0.42872101068496704, + "learning_rate": 3.822325287244975e-05, + "loss": 0.0268, + "step": 5970 + }, + { + "grad_norm": 0.5067957639694214, + "learning_rate": 3.8062622469526725e-05, + "loss": 0.0283, + "step": 5980 + }, + { + "grad_norm": 0.5961189866065979, + "learning_rate": 3.790212261185183e-05, + "loss": 0.0295, + "step": 5990 + }, + { + "grad_norm": 0.4409680962562561, + "learning_rate": 3.7741755054625794e-05, + "loss": 0.0303, + "step": 6000 + }, + { + "grad_norm": 0.34248724579811096, + "learning_rate": 3.758152155160255e-05, + "loss": 0.0281, + "step": 6010 + }, + { + "grad_norm": 0.4859353005886078, + "learning_rate": 3.742142385506999e-05, + "loss": 0.0295, + "step": 6020 + }, + { + "grad_norm": 0.47726792097091675, + "learning_rate": 3.72614637158309e-05, + "loss": 0.0253, + "step": 6030 + }, + { + "grad_norm": 0.38943901658058167, + "learning_rate": 3.710164288318371e-05, + "loss": 0.0312, + "step": 6040 + }, + { + "grad_norm": 0.36257413029670715, + "learning_rate": 3.694196310490345e-05, + "loss": 0.0249, + "step": 6050 + }, + { + "grad_norm": 0.5359441637992859, + "learning_rate": 3.678242612722259e-05, + "loss": 0.0281, + "step": 6060 + }, + { + "grad_norm": 0.4302363395690918, + "learning_rate": 3.6623033694811953e-05, + "loss": 0.0269, + "step": 6070 + }, + { + "grad_norm": 0.4097417891025543, + "learning_rate": 3.6463787550761665e-05, + "loss": 0.0298, + "step": 6080 + }, + { + "grad_norm": 0.3631576597690582, + "learning_rate": 3.630468943656202e-05, + "loss": 0.0272, + "step": 6090 + }, + { + "grad_norm": 0.3886636197566986, + "learning_rate": 3.6145741092084523e-05, + "loss": 0.0263, + "step": 6100 + }, + { + "grad_norm": 0.36417898535728455, + "learning_rate": 3.598694425556278e-05, + "loss": 0.0241, + "step": 6110 + }, + { + "grad_norm": 0.43107011914253235, + "learning_rate": 3.58283006635736e-05, + "loss": 0.0278, + "step": 6120 + }, + { + "grad_norm": 0.4701620638370514, + "learning_rate": 3.566981205101781e-05, + "loss": 0.0274, + "step": 6130 + }, + { + "grad_norm": 0.4055045247077942, + "learning_rate": 3.5511480151101556e-05, + "loss": 0.0266, + "step": 6140 + }, + { + "grad_norm": 0.46723777055740356, + "learning_rate": 3.5353306695317104e-05, + "loss": 0.024, + "step": 6150 + }, + { + "grad_norm": 0.4968148469924927, + "learning_rate": 3.519529341342402e-05, + "loss": 0.0322, + "step": 6160 + }, + { + "grad_norm": 0.36848726868629456, + "learning_rate": 3.503744203343026e-05, + "loss": 0.0293, + "step": 6170 + }, + { + "grad_norm": 0.4298679828643799, + "learning_rate": 3.487975428157318e-05, + "loss": 0.0291, + "step": 6180 + }, + { + "grad_norm": 0.48925650119781494, + "learning_rate": 3.472223188230083e-05, + "loss": 0.0294, + "step": 6190 + }, + { + "grad_norm": 0.4505534768104553, + "learning_rate": 3.4564876558252866e-05, + "loss": 0.0285, + "step": 6200 + }, + { + "grad_norm": 0.40965282917022705, + "learning_rate": 3.440769003024195e-05, + "loss": 0.0282, + "step": 6210 + }, + { + "grad_norm": 0.5326996445655823, + "learning_rate": 3.425067401723477e-05, + "loss": 0.0264, + "step": 6220 + }, + { + "grad_norm": 0.5371710658073425, + "learning_rate": 3.409383023633325e-05, + "loss": 0.0267, + "step": 6230 + }, + { + "grad_norm": 0.442018985748291, + "learning_rate": 3.3937160402755894e-05, + "loss": 0.026, + "step": 6240 + }, + { + "grad_norm": 0.5067367553710938, + "learning_rate": 3.378066622981885e-05, + "loss": 0.0273, + "step": 6250 + }, + { + "grad_norm": 0.349338561296463, + "learning_rate": 3.362434942891738e-05, + "loss": 0.0272, + "step": 6260 + }, + { + "grad_norm": 0.4490080177783966, + "learning_rate": 3.346821170950693e-05, + "loss": 0.0272, + "step": 6270 + }, + { + "grad_norm": 0.5187338590621948, + "learning_rate": 3.3312254779084585e-05, + "loss": 0.0268, + "step": 6280 + }, + { + "grad_norm": 0.49030300974845886, + "learning_rate": 3.315648034317039e-05, + "loss": 0.028, + "step": 6290 + }, + { + "grad_norm": 0.42636463046073914, + "learning_rate": 3.3000890105288564e-05, + "loss": 0.0241, + "step": 6300 + }, + { + "grad_norm": 0.42407524585723877, + "learning_rate": 3.284548576694908e-05, + "loss": 0.0255, + "step": 6310 + }, + { + "grad_norm": 0.37593695521354675, + "learning_rate": 3.2690269027628815e-05, + "loss": 0.0248, + "step": 6320 + }, + { + "grad_norm": 0.3526025414466858, + "learning_rate": 3.253524158475324e-05, + "loss": 0.0228, + "step": 6330 + }, + { + "grad_norm": 0.3837971091270447, + "learning_rate": 3.238040513367757e-05, + "loss": 0.0244, + "step": 6340 + }, + { + "grad_norm": 0.4616474509239197, + "learning_rate": 3.222576136766843e-05, + "loss": 0.022, + "step": 6350 + }, + { + "grad_norm": 0.4324616491794586, + "learning_rate": 3.2071311977885324e-05, + "loss": 0.0277, + "step": 6360 + }, + { + "grad_norm": 0.5130193829536438, + "learning_rate": 3.191705865336197e-05, + "loss": 0.0278, + "step": 6370 + }, + { + "grad_norm": 0.3285134732723236, + "learning_rate": 3.1763003080988075e-05, + "loss": 0.0254, + "step": 6380 + }, + { + "grad_norm": 0.49770838022232056, + "learning_rate": 3.160914694549063e-05, + "loss": 0.0255, + "step": 6390 + }, + { + "grad_norm": 0.47093814611434937, + "learning_rate": 3.145549192941573e-05, + "loss": 0.0261, + "step": 6400 + }, + { + "grad_norm": 0.5267713069915771, + "learning_rate": 3.130203971310999e-05, + "loss": 0.0253, + "step": 6410 + }, + { + "grad_norm": 0.42965826392173767, + "learning_rate": 3.114879197470225e-05, + "loss": 0.0271, + "step": 6420 + }, + { + "grad_norm": 0.48353004455566406, + "learning_rate": 3.0995750390085285e-05, + "loss": 0.0234, + "step": 6430 + }, + { + "grad_norm": 0.4051535129547119, + "learning_rate": 3.084291663289728e-05, + "loss": 0.0248, + "step": 6440 + }, + { + "grad_norm": 0.5486766695976257, + "learning_rate": 3.069029237450375e-05, + "loss": 0.0223, + "step": 6450 + }, + { + "grad_norm": 0.479905366897583, + "learning_rate": 3.053787928397911e-05, + "loss": 0.0231, + "step": 6460 + }, + { + "grad_norm": 0.48640313744544983, + "learning_rate": 3.0385679028088526e-05, + "loss": 0.0274, + "step": 6470 + }, + { + "grad_norm": 0.5368398427963257, + "learning_rate": 3.023369327126959e-05, + "loss": 0.0254, + "step": 6480 + }, + { + "grad_norm": 0.40994328260421753, + "learning_rate": 3.0081923675614198e-05, + "loss": 0.0246, + "step": 6490 + }, + { + "grad_norm": 0.46405959129333496, + "learning_rate": 2.993037190085034e-05, + "loss": 0.0244, + "step": 6500 + }, + { + "grad_norm": 0.5016579627990723, + "learning_rate": 2.977903960432392e-05, + "loss": 0.0272, + "step": 6510 + }, + { + "grad_norm": 0.45939400792121887, + "learning_rate": 2.9627928440980722e-05, + "loss": 0.0298, + "step": 6520 + }, + { + "grad_norm": 0.46316593885421753, + "learning_rate": 2.9477040063348183e-05, + "loss": 0.0275, + "step": 6530 + }, + { + "grad_norm": 0.4684191048145294, + "learning_rate": 2.9326376121517456e-05, + "loss": 0.0267, + "step": 6540 + }, + { + "grad_norm": 0.3520528972148895, + "learning_rate": 2.9175938263125236e-05, + "loss": 0.0248, + "step": 6550 + }, + { + "grad_norm": 0.41061222553253174, + "learning_rate": 2.9025728133335873e-05, + "loss": 0.0234, + "step": 6560 + }, + { + "grad_norm": 0.4299565851688385, + "learning_rate": 2.8875747374823288e-05, + "loss": 0.0276, + "step": 6570 + }, + { + "grad_norm": 0.5510832071304321, + "learning_rate": 2.872599762775298e-05, + "loss": 0.0323, + "step": 6580 + }, + { + "grad_norm": 0.38992759585380554, + "learning_rate": 2.857648052976425e-05, + "loss": 0.0262, + "step": 6590 + }, + { + "grad_norm": 0.42552703619003296, + "learning_rate": 2.8427197715952047e-05, + "loss": 0.0222, + "step": 6600 + }, + { + "grad_norm": 0.4920859932899475, + "learning_rate": 2.8278150818849393e-05, + "loss": 0.0239, + "step": 6610 + }, + { + "grad_norm": 0.3833717405796051, + "learning_rate": 2.812934146840922e-05, + "loss": 0.023, + "step": 6620 + }, + { + "grad_norm": 0.37559786438941956, + "learning_rate": 2.7980771291986764e-05, + "loss": 0.027, + "step": 6630 + }, + { + "grad_norm": 0.4519325792789459, + "learning_rate": 2.783244191432167e-05, + "loss": 0.0244, + "step": 6640 + }, + { + "grad_norm": 0.5249439477920532, + "learning_rate": 2.768435495752022e-05, + "loss": 0.0223, + "step": 6650 + }, + { + "grad_norm": 0.5495597720146179, + "learning_rate": 2.753651204103771e-05, + "loss": 0.0217, + "step": 6660 + }, + { + "grad_norm": 0.47961047291755676, + "learning_rate": 2.7388914781660523e-05, + "loss": 0.0216, + "step": 6670 + }, + { + "grad_norm": 0.4470529556274414, + "learning_rate": 2.7241564793488693e-05, + "loss": 0.0232, + "step": 6680 + }, + { + "grad_norm": 0.42591768503189087, + "learning_rate": 2.7094463687918037e-05, + "loss": 0.0255, + "step": 6690 + }, + { + "grad_norm": 0.507934033870697, + "learning_rate": 2.694761307362268e-05, + "loss": 0.0222, + "step": 6700 + }, + { + "grad_norm": 0.46535319089889526, + "learning_rate": 2.6801014556537467e-05, + "loss": 0.023, + "step": 6710 + }, + { + "grad_norm": 0.4839480519294739, + "learning_rate": 2.6654669739840243e-05, + "loss": 0.0215, + "step": 6720 + }, + { + "grad_norm": 0.45738548040390015, + "learning_rate": 2.650858022393451e-05, + "loss": 0.0246, + "step": 6730 + }, + { + "grad_norm": 0.43080613017082214, + "learning_rate": 2.6362747606431747e-05, + "loss": 0.0245, + "step": 6740 + }, + { + "grad_norm": 0.41232290863990784, + "learning_rate": 2.6217173482134172e-05, + "loss": 0.0207, + "step": 6750 + }, + { + "grad_norm": 0.5051129460334778, + "learning_rate": 2.6071859443017044e-05, + "loss": 0.0205, + "step": 6760 + }, + { + "grad_norm": 0.39422672986984253, + "learning_rate": 2.5926807078211414e-05, + "loss": 0.0189, + "step": 6770 + }, + { + "grad_norm": 0.4572248160839081, + "learning_rate": 2.5782017973986728e-05, + "loss": 0.0228, + "step": 6780 + }, + { + "grad_norm": 0.39599499106407166, + "learning_rate": 2.5637493713733374e-05, + "loss": 0.0237, + "step": 6790 + }, + { + "grad_norm": 0.5066280364990234, + "learning_rate": 2.549323587794559e-05, + "loss": 0.0271, + "step": 6800 + }, + { + "grad_norm": 0.4832857847213745, + "learning_rate": 2.5349246044203895e-05, + "loss": 0.0209, + "step": 6810 + }, + { + "grad_norm": 0.5020378232002258, + "learning_rate": 2.520552578715808e-05, + "loss": 0.0248, + "step": 6820 + }, + { + "grad_norm": 0.4275672733783722, + "learning_rate": 2.506207667850981e-05, + "loss": 0.0213, + "step": 6830 + }, + { + "grad_norm": 0.43956732749938965, + "learning_rate": 2.4918900286995555e-05, + "loss": 0.0222, + "step": 6840 + }, + { + "grad_norm": 0.4105176031589508, + "learning_rate": 2.4775998178369458e-05, + "loss": 0.0193, + "step": 6850 + }, + { + "grad_norm": 0.6171485185623169, + "learning_rate": 2.4633371915386017e-05, + "loss": 0.0272, + "step": 6860 + }, + { + "grad_norm": 0.46249234676361084, + "learning_rate": 2.4491023057783235e-05, + "loss": 0.0281, + "step": 6870 + }, + { + "grad_norm": 0.46037331223487854, + "learning_rate": 2.4348953162265375e-05, + "loss": 0.0276, + "step": 6880 + }, + { + "grad_norm": 0.3118123710155487, + "learning_rate": 2.420716378248607e-05, + "loss": 0.0241, + "step": 6890 + }, + { + "grad_norm": 0.5230092406272888, + "learning_rate": 2.4065656469031266e-05, + "loss": 0.0223, + "step": 6900 + }, + { + "grad_norm": 0.43938925862312317, + "learning_rate": 2.3924432769402268e-05, + "loss": 0.0229, + "step": 6910 + }, + { + "grad_norm": 0.32706278562545776, + "learning_rate": 2.3783494227998844e-05, + "loss": 0.0284, + "step": 6920 + }, + { + "grad_norm": 0.5035949349403381, + "learning_rate": 2.3642842386102264e-05, + "loss": 0.0207, + "step": 6930 + }, + { + "grad_norm": 0.36542919278144836, + "learning_rate": 2.3502478781858567e-05, + "loss": 0.0203, + "step": 6940 + }, + { + "grad_norm": 0.42827850580215454, + "learning_rate": 2.3362404950261628e-05, + "loss": 0.0239, + "step": 6950 + }, + { + "grad_norm": 0.42001834511756897, + "learning_rate": 2.3222622423136458e-05, + "loss": 0.0229, + "step": 6960 + }, + { + "grad_norm": 0.4608784317970276, + "learning_rate": 2.3083132729122332e-05, + "loss": 0.0241, + "step": 6970 + }, + { + "grad_norm": 0.4277324974536896, + "learning_rate": 2.294393739365621e-05, + "loss": 0.0205, + "step": 6980 + }, + { + "grad_norm": 0.38703712821006775, + "learning_rate": 2.2805037938956e-05, + "loss": 0.0247, + "step": 6990 + }, + { + "grad_norm": 0.4146232008934021, + "learning_rate": 2.266643588400386e-05, + "loss": 0.0217, + "step": 7000 + }, + { + "grad_norm": 0.4227413833141327, + "learning_rate": 2.252813274452969e-05, + "loss": 0.0214, + "step": 7010 + }, + { + "grad_norm": 0.3973981440067291, + "learning_rate": 2.2390130032994427e-05, + "loss": 0.0184, + "step": 7020 + }, + { + "grad_norm": 0.44454658031463623, + "learning_rate": 2.2252429258573633e-05, + "loss": 0.0212, + "step": 7030 + }, + { + "grad_norm": 0.5414749979972839, + "learning_rate": 2.2115031927140904e-05, + "loss": 0.0242, + "step": 7040 + }, + { + "grad_norm": 0.5495474934577942, + "learning_rate": 2.1977939541251463e-05, + "loss": 0.0271, + "step": 7050 + }, + { + "grad_norm": 0.39948463439941406, + "learning_rate": 2.1841153600125684e-05, + "loss": 0.0204, + "step": 7060 + }, + { + "grad_norm": 0.37931695580482483, + "learning_rate": 2.170467559963267e-05, + "loss": 0.0219, + "step": 7070 + }, + { + "grad_norm": 0.3962871730327606, + "learning_rate": 2.1568507032273982e-05, + "loss": 0.0191, + "step": 7080 + }, + { + "grad_norm": 0.3527192175388336, + "learning_rate": 2.1432649387167264e-05, + "loss": 0.0199, + "step": 7090 + }, + { + "grad_norm": 0.4015713036060333, + "learning_rate": 2.1297104150029973e-05, + "loss": 0.022, + "step": 7100 + }, + { + "grad_norm": 0.36687955260276794, + "learning_rate": 2.116187280316307e-05, + "loss": 0.0192, + "step": 7110 + }, + { + "grad_norm": 0.4049524664878845, + "learning_rate": 2.1026956825434908e-05, + "loss": 0.0224, + "step": 7120 + }, + { + "grad_norm": 0.3978455662727356, + "learning_rate": 2.0892357692265017e-05, + "loss": 0.02, + "step": 7130 + }, + { + "grad_norm": 0.4353801906108856, + "learning_rate": 2.0758076875607947e-05, + "loss": 0.0216, + "step": 7140 + }, + { + "grad_norm": 0.5277400016784668, + "learning_rate": 2.0624115843937207e-05, + "loss": 0.0236, + "step": 7150 + }, + { + "grad_norm": 0.45090940594673157, + "learning_rate": 2.0490476062229157e-05, + "loss": 0.0238, + "step": 7160 + }, + { + "grad_norm": 0.5202705264091492, + "learning_rate": 2.035715899194704e-05, + "loss": 0.0206, + "step": 7170 + }, + { + "grad_norm": 0.4476768970489502, + "learning_rate": 2.022416609102499e-05, + "loss": 0.0228, + "step": 7180 + }, + { + "grad_norm": 0.3838024139404297, + "learning_rate": 2.009149881385205e-05, + "loss": 0.0255, + "step": 7190 + }, + { + "grad_norm": 0.38800835609436035, + "learning_rate": 1.995915861125634e-05, + "loss": 0.0192, + "step": 7200 + }, + { + "grad_norm": 0.3800967335700989, + "learning_rate": 1.9827146930489065e-05, + "loss": 0.0166, + "step": 7210 + }, + { + "grad_norm": 0.40402209758758545, + "learning_rate": 1.9695465215208848e-05, + "loss": 0.0223, + "step": 7220 + }, + { + "grad_norm": 0.3277592062950134, + "learning_rate": 1.9564114905465813e-05, + "loss": 0.0184, + "step": 7230 + }, + { + "grad_norm": 0.3389796316623688, + "learning_rate": 1.9433097437685936e-05, + "loss": 0.0205, + "step": 7240 + }, + { + "grad_norm": 0.44041907787323, + "learning_rate": 1.930241424465521e-05, + "loss": 0.0175, + "step": 7250 + }, + { + "grad_norm": 0.40021464228630066, + "learning_rate": 1.9172066755504115e-05, + "loss": 0.0186, + "step": 7260 + }, + { + "grad_norm": 0.46017366647720337, + "learning_rate": 1.9042056395691914e-05, + "loss": 0.0195, + "step": 7270 + }, + { + "grad_norm": 0.3654099404811859, + "learning_rate": 1.8912384586991066e-05, + "loss": 0.018, + "step": 7280 + }, + { + "grad_norm": 0.4310210943222046, + "learning_rate": 1.8783052747471717e-05, + "loss": 0.0178, + "step": 7290 + }, + { + "grad_norm": 0.41026613116264343, + "learning_rate": 1.865406229148611e-05, + "loss": 0.0199, + "step": 7300 + }, + { + "grad_norm": 0.4320450723171234, + "learning_rate": 1.8525414629653233e-05, + "loss": 0.0238, + "step": 7310 + }, + { + "grad_norm": 0.40030914545059204, + "learning_rate": 1.8397111168843255e-05, + "loss": 0.0182, + "step": 7320 + }, + { + "grad_norm": 0.5210416913032532, + "learning_rate": 1.8269153312162323e-05, + "loss": 0.0229, + "step": 7330 + }, + { + "grad_norm": 0.39293450117111206, + "learning_rate": 1.8141542458937054e-05, + "loss": 0.0196, + "step": 7340 + }, + { + "grad_norm": 0.3522605001926422, + "learning_rate": 1.8014280004699268e-05, + "loss": 0.0169, + "step": 7350 + }, + { + "grad_norm": 0.5015451908111572, + "learning_rate": 1.788736734117078e-05, + "loss": 0.0176, + "step": 7360 + }, + { + "grad_norm": 0.3863304853439331, + "learning_rate": 1.7760805856248152e-05, + "loss": 0.0189, + "step": 7370 + }, + { + "grad_norm": 0.4394012987613678, + "learning_rate": 1.7634596933987518e-05, + "loss": 0.0191, + "step": 7380 + }, + { + "grad_norm": 0.49220702052116394, + "learning_rate": 1.7508741954589404e-05, + "loss": 0.0199, + "step": 7390 + }, + { + "grad_norm": 0.3319714665412903, + "learning_rate": 1.7383242294383717e-05, + "loss": 0.0192, + "step": 7400 + }, + { + "grad_norm": 0.39908361434936523, + "learning_rate": 1.7258099325814632e-05, + "loss": 0.017, + "step": 7410 + }, + { + "grad_norm": 0.3485678732395172, + "learning_rate": 1.7133314417425594e-05, + "loss": 0.0224, + "step": 7420 + }, + { + "grad_norm": 0.4289276897907257, + "learning_rate": 1.7008888933844408e-05, + "loss": 0.0222, + "step": 7430 + }, + { + "grad_norm": 0.39576002955436707, + "learning_rate": 1.6884824235768172e-05, + "loss": 0.0216, + "step": 7440 + }, + { + "grad_norm": 0.44759976863861084, + "learning_rate": 1.6761121679948592e-05, + "loss": 0.0179, + "step": 7450 + }, + { + "grad_norm": 0.3677750527858734, + "learning_rate": 1.663778261917695e-05, + "loss": 0.0211, + "step": 7460 + }, + { + "grad_norm": 0.431515634059906, + "learning_rate": 1.651480840226952e-05, + "loss": 0.0184, + "step": 7470 + }, + { + "grad_norm": 0.35328418016433716, + "learning_rate": 1.639220037405258e-05, + "loss": 0.0187, + "step": 7480 + }, + { + "grad_norm": 0.3630378544330597, + "learning_rate": 1.6269959875347906e-05, + "loss": 0.0207, + "step": 7490 + }, + { + "grad_norm": 0.3996571898460388, + "learning_rate": 1.614808824295802e-05, + "loss": 0.0171, + "step": 7500 + }, + { + "grad_norm": 0.39342862367630005, + "learning_rate": 1.602658680965152e-05, + "loss": 0.0175, + "step": 7510 + }, + { + "grad_norm": 0.3732790946960449, + "learning_rate": 1.5905456904148686e-05, + "loss": 0.021, + "step": 7520 + }, + { + "grad_norm": 0.40530213713645935, + "learning_rate": 1.57846998511067e-05, + "loss": 0.0199, + "step": 7530 + }, + { + "grad_norm": 0.3606318533420563, + "learning_rate": 1.566431697110538e-05, + "loss": 0.0177, + "step": 7540 + }, + { + "grad_norm": 0.4251377284526825, + "learning_rate": 1.554430958063259e-05, + "loss": 0.0166, + "step": 7550 + }, + { + "grad_norm": 0.31809544563293457, + "learning_rate": 1.5424678992069912e-05, + "loss": 0.0178, + "step": 7560 + }, + { + "grad_norm": 0.49615082144737244, + "learning_rate": 1.5305426513678362e-05, + "loss": 0.0219, + "step": 7570 + }, + { + "grad_norm": 0.40584394335746765, + "learning_rate": 1.518655344958388e-05, + "loss": 0.0209, + "step": 7580 + }, + { + "grad_norm": 0.4765651524066925, + "learning_rate": 1.5068061099763275e-05, + "loss": 0.0181, + "step": 7590 + }, + { + "grad_norm": 0.4534238576889038, + "learning_rate": 1.494995076002988e-05, + "loss": 0.0186, + "step": 7600 + }, + { + "grad_norm": 0.37714463472366333, + "learning_rate": 1.4832223722019456e-05, + "loss": 0.0162, + "step": 7610 + }, + { + "grad_norm": 0.290304034948349, + "learning_rate": 1.4714881273176035e-05, + "loss": 0.0174, + "step": 7620 + }, + { + "grad_norm": 0.37357765436172485, + "learning_rate": 1.4597924696737835e-05, + "loss": 0.0165, + "step": 7630 + }, + { + "grad_norm": 0.4153476357460022, + "learning_rate": 1.4481355271723252e-05, + "loss": 0.0167, + "step": 7640 + }, + { + "grad_norm": 0.4254070520401001, + "learning_rate": 1.4365174272916809e-05, + "loss": 0.0224, + "step": 7650 + }, + { + "grad_norm": 0.4493449032306671, + "learning_rate": 1.4249382970855319e-05, + "loss": 0.0184, + "step": 7660 + }, + { + "grad_norm": 0.35711076855659485, + "learning_rate": 1.4133982631813903e-05, + "loss": 0.017, + "step": 7670 + }, + { + "grad_norm": 0.32560962438583374, + "learning_rate": 1.4018974517792194e-05, + "loss": 0.0171, + "step": 7680 + }, + { + "grad_norm": 0.4263555407524109, + "learning_rate": 1.390435988650048e-05, + "loss": 0.0183, + "step": 7690 + }, + { + "grad_norm": 0.406827837228775, + "learning_rate": 1.3790139991346006e-05, + "loss": 0.0181, + "step": 7700 + }, + { + "grad_norm": 0.4140101969242096, + "learning_rate": 1.367631608141926e-05, + "loss": 0.0186, + "step": 7710 + }, + { + "grad_norm": 0.31141573190689087, + "learning_rate": 1.3562889401480278e-05, + "loss": 0.016, + "step": 7720 + }, + { + "grad_norm": 0.3120458126068115, + "learning_rate": 1.3449861191945074e-05, + "loss": 0.0194, + "step": 7730 + }, + { + "grad_norm": 0.35619691014289856, + "learning_rate": 1.3337232688872009e-05, + "loss": 0.0156, + "step": 7740 + }, + { + "grad_norm": 0.29900869727134705, + "learning_rate": 1.3225005123948364e-05, + "loss": 0.0148, + "step": 7750 + }, + { + "grad_norm": 0.3176744282245636, + "learning_rate": 1.311317972447681e-05, + "loss": 0.0158, + "step": 7760 + }, + { + "grad_norm": 0.32336723804473877, + "learning_rate": 1.3001757713361996e-05, + "loss": 0.0135, + "step": 7770 + }, + { + "grad_norm": 0.4444942772388458, + "learning_rate": 1.2890740309097204e-05, + "loss": 0.0193, + "step": 7780 + }, + { + "grad_norm": 0.37668877840042114, + "learning_rate": 1.2780128725750944e-05, + "loss": 0.0167, + "step": 7790 + }, + { + "grad_norm": 0.4025029242038727, + "learning_rate": 1.266992417295379e-05, + "loss": 0.0188, + "step": 7800 + }, + { + "grad_norm": 0.35728052258491516, + "learning_rate": 1.2560127855885073e-05, + "loss": 0.0197, + "step": 7810 + }, + { + "grad_norm": 0.42919692397117615, + "learning_rate": 1.2450740975259745e-05, + "loss": 0.0221, + "step": 7820 + }, + { + "grad_norm": 0.31981566548347473, + "learning_rate": 1.234176472731517e-05, + "loss": 0.0228, + "step": 7830 + }, + { + "grad_norm": 0.47693896293640137, + "learning_rate": 1.2233200303798158e-05, + "loss": 0.019, + "step": 7840 + }, + { + "grad_norm": 0.4110092520713806, + "learning_rate": 1.2125048891951846e-05, + "loss": 0.0156, + "step": 7850 + }, + { + "grad_norm": 0.3585768938064575, + "learning_rate": 1.2017311674502745e-05, + "loss": 0.0148, + "step": 7860 + }, + { + "grad_norm": 0.3853120803833008, + "learning_rate": 1.1909989829647822e-05, + "loss": 0.0193, + "step": 7870 + }, + { + "grad_norm": 0.38362768292427063, + "learning_rate": 1.1803084531041553e-05, + "loss": 0.0173, + "step": 7880 + }, + { + "grad_norm": 0.42706069350242615, + "learning_rate": 1.1696596947783162e-05, + "loss": 0.0156, + "step": 7890 + }, + { + "grad_norm": 0.3532147705554962, + "learning_rate": 1.1590528244403803e-05, + "loss": 0.0172, + "step": 7900 + }, + { + "grad_norm": 0.3714560270309448, + "learning_rate": 1.148487958085382e-05, + "loss": 0.0188, + "step": 7910 + }, + { + "grad_norm": 0.3001372814178467, + "learning_rate": 1.1379652112490086e-05, + "loss": 0.0152, + "step": 7920 + }, + { + "grad_norm": 0.32501310110092163, + "learning_rate": 1.1274846990063315e-05, + "loss": 0.0187, + "step": 7930 + }, + { + "grad_norm": 0.4298228621482849, + "learning_rate": 1.117046535970554e-05, + "loss": 0.0154, + "step": 7940 + }, + { + "grad_norm": 0.3164469003677368, + "learning_rate": 1.106650836291755e-05, + "loss": 0.0215, + "step": 7950 + }, + { + "grad_norm": 0.45364123582839966, + "learning_rate": 1.0962977136556418e-05, + "loss": 0.0217, + "step": 7960 + }, + { + "grad_norm": 0.37747427821159363, + "learning_rate": 1.0859872812823024e-05, + "loss": 0.019, + "step": 7970 + }, + { + "grad_norm": 0.3764651417732239, + "learning_rate": 1.0757196519249747e-05, + "loss": 0.0248, + "step": 7980 + }, + { + "grad_norm": 0.5184255242347717, + "learning_rate": 1.0654949378688077e-05, + "loss": 0.0196, + "step": 7990 + }, + { + "grad_norm": 0.3421761393547058, + "learning_rate": 1.0553132509296376e-05, + "loss": 0.0168, + "step": 8000 + }, + { + "grad_norm": 0.3845522999763489, + "learning_rate": 1.0451747024527613e-05, + "loss": 0.0186, + "step": 8010 + }, + { + "grad_norm": 0.3916364908218384, + "learning_rate": 1.0350794033117189e-05, + "loss": 0.0162, + "step": 8020 + }, + { + "grad_norm": 0.313238263130188, + "learning_rate": 1.0250274639070856e-05, + "loss": 0.0164, + "step": 8030 + }, + { + "grad_norm": 0.28323641419410706, + "learning_rate": 1.0150189941652599e-05, + "loss": 0.0147, + "step": 8040 + }, + { + "grad_norm": 0.37788036465644836, + "learning_rate": 1.0050541035372635e-05, + "loss": 0.0146, + "step": 8050 + }, + { + "grad_norm": 0.4246368706226349, + "learning_rate": 9.951329009975458e-06, + "loss": 0.0189, + "step": 8060 + }, + { + "grad_norm": 0.365826815366745, + "learning_rate": 9.852554950427845e-06, + "loss": 0.0185, + "step": 8070 + }, + { + "grad_norm": 0.39302152395248413, + "learning_rate": 9.754219936907105e-06, + "loss": 0.0163, + "step": 8080 + }, + { + "grad_norm": 0.3342374265193939, + "learning_rate": 9.656325044789194e-06, + "loss": 0.0152, + "step": 8090 + }, + { + "grad_norm": 0.39748474955558777, + "learning_rate": 9.55887134463697e-06, + "loss": 0.0169, + "step": 8100 + }, + { + "grad_norm": 0.3542501628398895, + "learning_rate": 9.461859902188475e-06, + "loss": 0.0166, + "step": 8110 + }, + { + "grad_norm": 0.2805088758468628, + "learning_rate": 9.365291778345303e-06, + "loss": 0.0162, + "step": 8120 + }, + { + "grad_norm": 0.2923654317855835, + "learning_rate": 9.269168029160991e-06, + "loss": 0.0159, + "step": 8130 + }, + { + "grad_norm": 0.36373454332351685, + "learning_rate": 9.173489705829447e-06, + "loss": 0.0178, + "step": 8140 + }, + { + "grad_norm": 0.3142955005168915, + "learning_rate": 9.078257854673516e-06, + "loss": 0.0166, + "step": 8150 + }, + { + "grad_norm": 0.34815073013305664, + "learning_rate": 8.983473517133429e-06, + "loss": 0.016, + "step": 8160 + }, + { + "grad_norm": 0.40496838092803955, + "learning_rate": 8.889137729755537e-06, + "loss": 0.0161, + "step": 8170 + }, + { + "grad_norm": 0.3643417954444885, + "learning_rate": 8.79525152418087e-06, + "loss": 0.0179, + "step": 8180 + }, + { + "grad_norm": 0.35140982270240784, + "learning_rate": 8.701815927133961e-06, + "loss": 0.0142, + "step": 8190 + }, + { + "grad_norm": 0.35495054721832275, + "learning_rate": 8.608831960411534e-06, + "loss": 0.0157, + "step": 8200 + }, + { + "grad_norm": 0.3329131007194519, + "learning_rate": 8.516300640871321e-06, + "loss": 0.0153, + "step": 8210 + }, + { + "grad_norm": 0.39769527316093445, + "learning_rate": 8.424222980421038e-06, + "loss": 0.0174, + "step": 8220 + }, + { + "grad_norm": 0.3536461293697357, + "learning_rate": 8.332599986007184e-06, + "loss": 0.0165, + "step": 8230 + }, + { + "grad_norm": 0.3056160509586334, + "learning_rate": 8.241432659604203e-06, + "loss": 0.014, + "step": 8240 + }, + { + "grad_norm": 0.39745578169822693, + "learning_rate": 8.150721998203331e-06, + "loss": 0.0138, + "step": 8250 + }, + { + "grad_norm": 0.38860023021698, + "learning_rate": 8.06046899380184e-06, + "loss": 0.0147, + "step": 8260 + }, + { + "grad_norm": 0.3471664488315582, + "learning_rate": 7.970674633392133e-06, + "loss": 0.0218, + "step": 8270 + }, + { + "grad_norm": 0.24197325110435486, + "learning_rate": 7.881339898950924e-06, + "loss": 0.0158, + "step": 8280 + }, + { + "grad_norm": 0.2840465307235718, + "learning_rate": 7.792465767428597e-06, + "loss": 0.0163, + "step": 8290 + }, + { + "grad_norm": 0.2441285252571106, + "learning_rate": 7.704053210738376e-06, + "loss": 0.0141, + "step": 8300 + }, + { + "grad_norm": 0.394169420003891, + "learning_rate": 7.6161031957458494e-06, + "loss": 0.0185, + "step": 8310 + }, + { + "grad_norm": 0.4517776370048523, + "learning_rate": 7.5286166842582605e-06, + "loss": 0.0165, + "step": 8320 + }, + { + "grad_norm": 0.3767309784889221, + "learning_rate": 7.4415946330140814e-06, + "loss": 0.016, + "step": 8330 + }, + { + "grad_norm": 0.43256697058677673, + "learning_rate": 7.3550379936725644e-06, + "loss": 0.016, + "step": 8340 + }, + { + "grad_norm": 0.3534623086452484, + "learning_rate": 7.2689477128032035e-06, + "loss": 0.0178, + "step": 8350 + }, + { + "grad_norm": 0.2946368157863617, + "learning_rate": 7.183324731875551e-06, + "loss": 0.0142, + "step": 8360 + }, + { + "grad_norm": 0.3349359333515167, + "learning_rate": 7.098169987248782e-06, + "loss": 0.0143, + "step": 8370 + }, + { + "grad_norm": 0.29287275671958923, + "learning_rate": 7.013484410161553e-06, + "loss": 0.0151, + "step": 8380 + }, + { + "grad_norm": 0.2697104513645172, + "learning_rate": 6.92926892672176e-06, + "loss": 0.0156, + "step": 8390 + }, + { + "grad_norm": 0.22261130809783936, + "learning_rate": 6.845524457896446e-06, + "loss": 0.0138, + "step": 8400 + }, + { + "grad_norm": 0.29333293437957764, + "learning_rate": 6.7622519195017165e-06, + "loss": 0.0147, + "step": 8410 + }, + { + "grad_norm": 0.35222089290618896, + "learning_rate": 6.679452222192684e-06, + "loss": 0.0141, + "step": 8420 + }, + { + "grad_norm": 0.36629948019981384, + "learning_rate": 6.597126271453579e-06, + "loss": 0.0164, + "step": 8430 + }, + { + "grad_norm": 0.3296319842338562, + "learning_rate": 6.51527496758782e-06, + "loss": 0.0148, + "step": 8440 + }, + { + "grad_norm": 0.2617538571357727, + "learning_rate": 6.433899205708155e-06, + "loss": 0.014, + "step": 8450 + }, + { + "grad_norm": 0.36103323101997375, + "learning_rate": 6.352999875726856e-06, + "loss": 0.0132, + "step": 8460 + }, + { + "grad_norm": 0.27877333760261536, + "learning_rate": 6.272577862346052e-06, + "loss": 0.0156, + "step": 8470 + }, + { + "grad_norm": 0.41543734073638916, + "learning_rate": 6.192634045047996e-06, + "loss": 0.0226, + "step": 8480 + }, + { + "grad_norm": 0.2565414011478424, + "learning_rate": 6.113169298085458e-06, + "loss": 0.0182, + "step": 8490 + }, + { + "grad_norm": 0.3711279332637787, + "learning_rate": 6.034184490472195e-06, + "loss": 0.015, + "step": 8500 + }, + { + "grad_norm": 0.32403960824012756, + "learning_rate": 5.955680485973386e-06, + "loss": 0.0154, + "step": 8510 + }, + { + "grad_norm": 0.324998676776886, + "learning_rate": 5.877658143096265e-06, + "loss": 0.0145, + "step": 8520 + }, + { + "grad_norm": 0.2539767920970917, + "learning_rate": 5.800118315080661e-06, + "loss": 0.014, + "step": 8530 + }, + { + "grad_norm": 0.30353474617004395, + "learning_rate": 5.723061849889716e-06, + "loss": 0.0159, + "step": 8540 + }, + { + "grad_norm": 0.2422187477350235, + "learning_rate": 5.646489590200604e-06, + "loss": 0.0164, + "step": 8550 + }, + { + "grad_norm": 0.35298141837120056, + "learning_rate": 5.570402373395256e-06, + "loss": 0.0146, + "step": 8560 + }, + { + "grad_norm": 0.2959458827972412, + "learning_rate": 5.494801031551305e-06, + "loss": 0.0169, + "step": 8570 + }, + { + "grad_norm": 0.26740020513534546, + "learning_rate": 5.41968639143291e-06, + "loss": 0.0132, + "step": 8580 + }, + { + "grad_norm": 0.33075201511383057, + "learning_rate": 5.345059274481751e-06, + "loss": 0.0135, + "step": 8590 + }, + { + "grad_norm": 0.31155475974082947, + "learning_rate": 5.270920496808002e-06, + "loss": 0.0155, + "step": 8600 + }, + { + "grad_norm": 0.3762166500091553, + "learning_rate": 5.1972708691814695e-06, + "loss": 0.0186, + "step": 8610 + }, + { + "grad_norm": 0.40252411365509033, + "learning_rate": 5.124111197022674e-06, + "loss": 0.0175, + "step": 8620 + }, + { + "grad_norm": 0.29389381408691406, + "learning_rate": 5.051442280394081e-06, + "loss": 0.0135, + "step": 8630 + }, + { + "grad_norm": 0.37129899859428406, + "learning_rate": 4.979264913991322e-06, + "loss": 0.0139, + "step": 8640 + }, + { + "grad_norm": 0.29435938596725464, + "learning_rate": 4.907579887134489e-06, + "loss": 0.0196, + "step": 8650 + }, + { + "grad_norm": 0.4199511706829071, + "learning_rate": 4.836387983759572e-06, + "loss": 0.0144, + "step": 8660 + }, + { + "grad_norm": 0.35561540722846985, + "learning_rate": 4.765689982409816e-06, + "loss": 0.0172, + "step": 8670 + }, + { + "grad_norm": 0.3386131525039673, + "learning_rate": 4.695486656227233e-06, + "loss": 0.0192, + "step": 8680 + }, + { + "grad_norm": 0.35949355363845825, + "learning_rate": 4.625778772944156e-06, + "loss": 0.015, + "step": 8690 + }, + { + "grad_norm": 0.3971130847930908, + "learning_rate": 4.556567094874825e-06, + "loss": 0.0163, + "step": 8700 + }, + { + "grad_norm": 0.32525068521499634, + "learning_rate": 4.487852378907059e-06, + "loss": 0.0128, + "step": 8710 + }, + { + "grad_norm": 0.2896662652492523, + "learning_rate": 4.419635376493986e-06, + "loss": 0.0149, + "step": 8720 + }, + { + "grad_norm": 0.3105625808238983, + "learning_rate": 4.351916833645825e-06, + "loss": 0.0159, + "step": 8730 + }, + { + "grad_norm": 0.40515458583831787, + "learning_rate": 4.284697490921691e-06, + "loss": 0.0154, + "step": 8740 + }, + { + "grad_norm": 0.3142366111278534, + "learning_rate": 4.2179780834215585e-06, + "loss": 0.0156, + "step": 8750 + }, + { + "grad_norm": 0.3781340718269348, + "learning_rate": 4.151759340778178e-06, + "loss": 0.0159, + "step": 8760 + }, + { + "grad_norm": 0.23719792068004608, + "learning_rate": 4.086041987149109e-06, + "loss": 0.0132, + "step": 8770 + }, + { + "grad_norm": 0.3757002055644989, + "learning_rate": 4.020826741208811e-06, + "loss": 0.0186, + "step": 8780 + }, + { + "grad_norm": 0.2862277328968048, + "learning_rate": 3.956114316140746e-06, + "loss": 0.0148, + "step": 8790 + }, + { + "grad_norm": 0.2751687467098236, + "learning_rate": 3.891905419629643e-06, + "loss": 0.0141, + "step": 8800 + }, + { + "grad_norm": 0.2684212028980255, + "learning_rate": 3.8282007538536946e-06, + "loss": 0.0182, + "step": 8810 + }, + { + "grad_norm": 0.38802894949913025, + "learning_rate": 3.7650010154769265e-06, + "loss": 0.0156, + "step": 8820 + }, + { + "grad_norm": 0.2776685655117035, + "learning_rate": 3.7023068956415608e-06, + "loss": 0.0149, + "step": 8830 + }, + { + "grad_norm": 0.3029594421386719, + "learning_rate": 3.6401190799604303e-06, + "loss": 0.0134, + "step": 8840 + }, + { + "grad_norm": 0.2905641794204712, + "learning_rate": 3.578438248509536e-06, + "loss": 0.0156, + "step": 8850 + }, + { + "grad_norm": 0.3464786112308502, + "learning_rate": 3.5172650758205583e-06, + "loss": 0.0159, + "step": 8860 + }, + { + "grad_norm": 0.24294251203536987, + "learning_rate": 3.45660023087353e-06, + "loss": 0.0122, + "step": 8870 + }, + { + "grad_norm": 0.25388672947883606, + "learning_rate": 3.3964443770894528e-06, + "loss": 0.016, + "step": 8880 + }, + { + "grad_norm": 0.24941787123680115, + "learning_rate": 3.3367981723231245e-06, + "loss": 0.0133, + "step": 8890 + }, + { + "grad_norm": 0.3475058972835541, + "learning_rate": 3.2776622688558746e-06, + "loss": 0.0175, + "step": 8900 + }, + { + "grad_norm": 0.2986414134502411, + "learning_rate": 3.2190373133884677e-06, + "loss": 0.0113, + "step": 8910 + }, + { + "grad_norm": 0.27067404985427856, + "learning_rate": 3.1609239470340446e-06, + "loss": 0.0126, + "step": 8920 + }, + { + "grad_norm": 0.18267281353473663, + "learning_rate": 3.1033228053110373e-06, + "loss": 0.0123, + "step": 8930 + }, + { + "grad_norm": 0.213578999042511, + "learning_rate": 3.0462345181363314e-06, + "loss": 0.0139, + "step": 8940 + }, + { + "grad_norm": 0.3505679666996002, + "learning_rate": 2.9896597098182654e-06, + "loss": 0.0173, + "step": 8950 + }, + { + "grad_norm": 0.3307925760746002, + "learning_rate": 2.933598999049891e-06, + "loss": 0.0163, + "step": 8960 + }, + { + "grad_norm": 0.24764008820056915, + "learning_rate": 2.8780529989021697e-06, + "loss": 0.013, + "step": 8970 + }, + { + "grad_norm": 0.2879706919193268, + "learning_rate": 2.823022316817242e-06, + "loss": 0.0167, + "step": 8980 + }, + { + "grad_norm": 0.22704508900642395, + "learning_rate": 2.7685075546018456e-06, + "loss": 0.0157, + "step": 8990 + }, + { + "grad_norm": 0.2594397962093353, + "learning_rate": 2.7145093084206598e-06, + "loss": 0.0188, + "step": 9000 + }, + { + "grad_norm": 0.23858588933944702, + "learning_rate": 2.661028168789892e-06, + "loss": 0.0141, + "step": 9010 + }, + { + "grad_norm": 0.29953068494796753, + "learning_rate": 2.6080647205706855e-06, + "loss": 0.0126, + "step": 9020 + }, + { + "grad_norm": 0.3014177978038788, + "learning_rate": 2.555619542962834e-06, + "loss": 0.0145, + "step": 9030 + }, + { + "grad_norm": 0.2251807302236557, + "learning_rate": 2.503693209498409e-06, + "loss": 0.0128, + "step": 9040 + }, + { + "grad_norm": 0.23174887895584106, + "learning_rate": 2.452286288035449e-06, + "loss": 0.0134, + "step": 9050 + }, + { + "grad_norm": 0.21669137477874756, + "learning_rate": 2.4013993407518363e-06, + "loss": 0.0149, + "step": 9060 + }, + { + "grad_norm": 0.29240232706069946, + "learning_rate": 2.351032924139063e-06, + "loss": 0.0136, + "step": 9070 + }, + { + "grad_norm": 0.25830525159835815, + "learning_rate": 2.30118758899619e-06, + "loss": 0.016, + "step": 9080 + }, + { + "grad_norm": 0.29298263788223267, + "learning_rate": 2.2518638804238157e-06, + "loss": 0.0143, + "step": 9090 + }, + { + "grad_norm": 0.28530094027519226, + "learning_rate": 2.203062337818118e-06, + "loss": 0.0167, + "step": 9100 + }, + { + "grad_norm": 0.30774784088134766, + "learning_rate": 2.1547834948649483e-06, + "loss": 0.0167, + "step": 9110 + }, + { + "grad_norm": 0.314654141664505, + "learning_rate": 2.1070278795340017e-06, + "loss": 0.0157, + "step": 9120 + }, + { + "grad_norm": 0.28017422556877136, + "learning_rate": 2.059796014073029e-06, + "loss": 0.0122, + "step": 9130 + }, + { + "grad_norm": 0.4184361398220062, + "learning_rate": 2.01308841500214e-06, + "loss": 0.0158, + "step": 9140 + }, + { + "grad_norm": 0.30968302488327026, + "learning_rate": 1.9669055931081704e-06, + "loss": 0.0191, + "step": 9150 + }, + { + "grad_norm": 0.24011468887329102, + "learning_rate": 1.9212480534390507e-06, + "loss": 0.0154, + "step": 9160 + }, + { + "grad_norm": 0.46455803513526917, + "learning_rate": 1.8761162952983246e-06, + "loss": 0.0199, + "step": 9170 + }, + { + "grad_norm": 0.2779623568058014, + "learning_rate": 1.8315108122396618e-06, + "loss": 0.017, + "step": 9180 + }, + { + "grad_norm": 0.2468317598104477, + "learning_rate": 1.787432092061475e-06, + "loss": 0.0123, + "step": 9190 + }, + { + "grad_norm": 0.23731543123722076, + "learning_rate": 1.743880616801602e-06, + "loss": 0.0158, + "step": 9200 + }, + { + "grad_norm": 0.3605269193649292, + "learning_rate": 1.7008568627319865e-06, + "loss": 0.0169, + "step": 9210 + }, + { + "grad_norm": 0.2651381492614746, + "learning_rate": 1.6583613003535226e-06, + "loss": 0.0168, + "step": 9220 + }, + { + "grad_norm": 0.2924129366874695, + "learning_rate": 1.6163943943908522e-06, + "loss": 0.0135, + "step": 9230 + }, + { + "grad_norm": 0.2569584846496582, + "learning_rate": 1.5749566037873476e-06, + "loss": 0.0137, + "step": 9240 + }, + { + "grad_norm": 0.32348793745040894, + "learning_rate": 1.5340483817000428e-06, + "loss": 0.0169, + "step": 9250 + }, + { + "grad_norm": 0.2810918092727661, + "learning_rate": 1.4936701754947101e-06, + "loss": 0.016, + "step": 9260 + }, + { + "grad_norm": 0.22482751309871674, + "learning_rate": 1.4538224267409361e-06, + "loss": 0.0136, + "step": 9270 + }, + { + "grad_norm": 0.34207969903945923, + "learning_rate": 1.414505571207314e-06, + "loss": 0.014, + "step": 9280 + }, + { + "grad_norm": 0.1883111298084259, + "learning_rate": 1.3757200388566816e-06, + "loss": 0.0148, + "step": 9290 + }, + { + "grad_norm": 0.2550380229949951, + "learning_rate": 1.3374662538414074e-06, + "loss": 0.0158, + "step": 9300 + }, + { + "grad_norm": 0.19865429401397705, + "learning_rate": 1.2997446344987617e-06, + "loss": 0.0152, + "step": 9310 + }, + { + "grad_norm": 0.2502734363079071, + "learning_rate": 1.262555593346315e-06, + "loss": 0.0132, + "step": 9320 + }, + { + "grad_norm": 0.2847469747066498, + "learning_rate": 1.2258995370774685e-06, + "loss": 0.0151, + "step": 9330 + }, + { + "grad_norm": 0.252002477645874, + "learning_rate": 1.1897768665569798e-06, + "loss": 0.0134, + "step": 9340 + }, + { + "grad_norm": 0.22425878047943115, + "learning_rate": 1.1541879768165954e-06, + "loss": 0.0128, + "step": 9350 + }, + { + "grad_norm": 0.31090983748435974, + "learning_rate": 1.1191332570507085e-06, + "loss": 0.0167, + "step": 9360 + }, + { + "grad_norm": 0.31644028425216675, + "learning_rate": 1.0846130906121132e-06, + "loss": 0.0188, + "step": 9370 + }, + { + "grad_norm": 0.2684969902038574, + "learning_rate": 1.0506278550078131e-06, + "loss": 0.0172, + "step": 9380 + }, + { + "grad_norm": 0.3127434253692627, + "learning_rate": 1.0171779218949185e-06, + "loss": 0.0168, + "step": 9390 + }, + { + "grad_norm": 0.18024255335330963, + "learning_rate": 9.842636570765174e-07, + "loss": 0.0127, + "step": 9400 + }, + { + "grad_norm": 0.22021129727363586, + "learning_rate": 9.518854204977612e-07, + "loss": 0.013, + "step": 9410 + }, + { + "grad_norm": 0.21553249657154083, + "learning_rate": 9.200435662418349e-07, + "loss": 0.0149, + "step": 9420 + }, + { + "grad_norm": 0.27088266611099243, + "learning_rate": 8.887384425261658e-07, + "loss": 0.0136, + "step": 9430 + }, + { + "grad_norm": 0.2857940196990967, + "learning_rate": 8.579703916985648e-07, + "loss": 0.0206, + "step": 9440 + }, + { + "grad_norm": 0.20183144509792328, + "learning_rate": 8.277397502335194e-07, + "loss": 0.0135, + "step": 9450 + }, + { + "grad_norm": 0.2638828158378601, + "learning_rate": 7.980468487284675e-07, + "loss": 0.0166, + "step": 9460 + }, + { + "grad_norm": 0.24388478696346283, + "learning_rate": 7.688920119002297e-07, + "loss": 0.0182, + "step": 9470 + }, + { + "grad_norm": 0.2569715082645416, + "learning_rate": 7.402755585814269e-07, + "loss": 0.0159, + "step": 9480 + }, + { + "grad_norm": 0.21388714015483856, + "learning_rate": 7.121978017170073e-07, + "loss": 0.0165, + "step": 9490 + }, + { + "grad_norm": 0.3321588933467865, + "learning_rate": 6.846590483608306e-07, + "loss": 0.0166, + "step": 9500 + }, + { + "grad_norm": 0.21789860725402832, + "learning_rate": 6.576595996722834e-07, + "loss": 0.0156, + "step": 9510 + }, + { + "grad_norm": 0.26565298438072205, + "learning_rate": 6.311997509130141e-07, + "loss": 0.0169, + "step": 9520 + }, + { + "grad_norm": 0.2036224603652954, + "learning_rate": 6.052797914436803e-07, + "loss": 0.0152, + "step": 9530 + }, + { + "grad_norm": 0.35222306847572327, + "learning_rate": 5.799000047208181e-07, + "loss": 0.0178, + "step": 9540 + }, + { + "grad_norm": 0.2662196159362793, + "learning_rate": 5.550606682937054e-07, + "loss": 0.0156, + "step": 9550 + }, + { + "grad_norm": 0.3069530427455902, + "learning_rate": 5.307620538013481e-07, + "loss": 0.0166, + "step": 9560 + }, + { + "grad_norm": 0.22157780826091766, + "learning_rate": 5.070044269694874e-07, + "loss": 0.0149, + "step": 9570 + }, + { + "grad_norm": 0.2512767016887665, + "learning_rate": 4.837880476077417e-07, + "loss": 0.017, + "step": 9580 + }, + { + "grad_norm": 0.25105831027030945, + "learning_rate": 4.6111316960670835e-07, + "loss": 0.0115, + "step": 9590 + }, + { + "grad_norm": 0.21751675009727478, + "learning_rate": 4.389800409352218e-07, + "loss": 0.0145, + "step": 9600 + }, + { + "grad_norm": 0.2965623438358307, + "learning_rate": 4.173889036376277e-07, + "loss": 0.017, + "step": 9610 + }, + { + "grad_norm": 0.31083691120147705, + "learning_rate": 3.963399938311463e-07, + "loss": 0.016, + "step": 9620 + }, + { + "grad_norm": 0.2296469807624817, + "learning_rate": 3.7583354170328545e-07, + "loss": 0.0165, + "step": 9630 + }, + { + "grad_norm": 0.16305774450302124, + "learning_rate": 3.558697715093207e-07, + "loss": 0.0121, + "step": 9640 + }, + { + "grad_norm": 0.281929612159729, + "learning_rate": 3.3644890156983576e-07, + "loss": 0.0153, + "step": 9650 + }, + { + "grad_norm": 0.2220996767282486, + "learning_rate": 3.175711442683638e-07, + "loss": 0.0123, + "step": 9660 + }, + { + "grad_norm": 0.16283515095710754, + "learning_rate": 2.9923670604902197e-07, + "loss": 0.015, + "step": 9670 + }, + { + "grad_norm": 0.25990134477615356, + "learning_rate": 2.814457874143028e-07, + "loss": 0.0165, + "step": 9680 + }, + { + "grad_norm": 0.20940668880939484, + "learning_rate": 2.641985829228366e-07, + "loss": 0.0127, + "step": 9690 + }, + { + "grad_norm": 0.17900049686431885, + "learning_rate": 2.474952811872877e-07, + "loss": 0.0125, + "step": 9700 + }, + { + "grad_norm": 0.17908349633216858, + "learning_rate": 2.3133606487228397e-07, + "loss": 0.0155, + "step": 9710 + }, + { + "grad_norm": 0.21241281926631927, + "learning_rate": 2.157211106924295e-07, + "loss": 0.017, + "step": 9720 + }, + { + "grad_norm": 0.24554672837257385, + "learning_rate": 2.006505894103672e-07, + "loss": 0.0152, + "step": 9730 + }, + { + "grad_norm": 0.17558197677135468, + "learning_rate": 1.8612466583489696e-07, + "loss": 0.015, + "step": 9740 + }, + { + "grad_norm": 0.24153850972652435, + "learning_rate": 1.7214349881918834e-07, + "loss": 0.0145, + "step": 9750 + }, + { + "grad_norm": 0.21432065963745117, + "learning_rate": 1.5870724125904845e-07, + "loss": 0.0149, + "step": 9760 + }, + { + "grad_norm": 0.26208266615867615, + "learning_rate": 1.4581604009124006e-07, + "loss": 0.0133, + "step": 9770 + }, + { + "grad_norm": 0.2023925483226776, + "learning_rate": 1.334700362918717e-07, + "loss": 0.0135, + "step": 9780 + }, + { + "grad_norm": 0.3117850422859192, + "learning_rate": 1.2166936487486015e-07, + "loss": 0.0167, + "step": 9790 + }, + { + "grad_norm": 0.26498812437057495, + "learning_rate": 1.1041415489045914e-07, + "loss": 0.0143, + "step": 9800 + }, + { + "grad_norm": 0.2646230161190033, + "learning_rate": 9.970452942384412e-08, + "loss": 0.0159, + "step": 9810 + }, + { + "grad_norm": 0.23477981984615326, + "learning_rate": 8.954060559375754e-08, + "loss": 0.0119, + "step": 9820 + }, + { + "grad_norm": 0.17755717039108276, + "learning_rate": 7.99224945512489e-08, + "loss": 0.0172, + "step": 9830 + }, + { + "grad_norm": 0.23726916313171387, + "learning_rate": 7.085030147843675e-08, + "loss": 0.0169, + "step": 9840 + }, + { + "grad_norm": 0.19380071759223938, + "learning_rate": 6.232412558736523e-08, + "loss": 0.0174, + "step": 9850 + }, + { + "grad_norm": 0.2483823150396347, + "learning_rate": 5.434406011893822e-08, + "loss": 0.0175, + "step": 9860 + }, + { + "grad_norm": 0.28488603234291077, + "learning_rate": 4.6910192341864664e-08, + "loss": 0.0124, + "step": 9870 + }, + { + "grad_norm": 0.2530451714992523, + "learning_rate": 4.0022603551737035e-08, + "loss": 0.0159, + "step": 9880 + }, + { + "grad_norm": 0.24549846351146698, + "learning_rate": 3.3681369070120985e-08, + "loss": 0.0173, + "step": 9890 + }, + { + "grad_norm": 0.26719021797180176, + "learning_rate": 2.7886558243744866e-08, + "loss": 0.016, + "step": 9900 + }, + { + "grad_norm": 0.2937699556350708, + "learning_rate": 2.2638234443722596e-08, + "loss": 0.017, + "step": 9910 + }, + { + "grad_norm": 0.3065221309661865, + "learning_rate": 1.7936455064887504e-08, + "loss": 0.0173, + "step": 9920 + }, + { + "grad_norm": 0.15951815247535706, + "learning_rate": 1.378127152514841e-08, + "loss": 0.0117, + "step": 9930 + }, + { + "grad_norm": 0.19330716133117676, + "learning_rate": 1.0172729264917857e-08, + "loss": 0.0173, + "step": 9940 + }, + { + "grad_norm": 0.31932997703552246, + "learning_rate": 7.1108677466458215e-09, + "loss": 0.0148, + "step": 9950 + }, + { + "grad_norm": 0.19208914041519165, + "learning_rate": 4.595720454353414e-09, + "loss": 0.0136, + "step": 9960 + }, + { + "grad_norm": 0.2848373353481293, + "learning_rate": 2.627314893294264e-09, + "loss": 0.0129, + "step": 9970 + }, + { + "grad_norm": 0.1740635484457016, + "learning_rate": 1.2056725896270048e-09, + "loss": 0.0142, + "step": 9980 + }, + { + "grad_norm": 0.22473189234733582, + "learning_rate": 3.308090902098826e-10, + "loss": 0.0157, + "step": 9990 + }, + { + "grad_norm": 0.22421416640281677, + "learning_rate": 2.7339624120159555e-12, + "loss": 0.0158, + "step": 10000 + } + ], + "logging_steps": 10, + "max_steps": 10000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 3000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}